Skip to content

Instantly share code, notes, and snippets.

@raman91

raman91/install_R_packages.R

Last active January 2, 2024 08:10
Show Gist options
  • Save raman91/5171381ae540bc7f58f43f79af2c7d6d to your computer and use it in GitHub Desktop.
Save raman91/5171381ae540bc7f58f43f79af2c7d6d to your computer and use it in GitHub Desktop.
Download ZIP
ezSingleCell
Raw
install_R_packages.R
options(repos = BiocManager::repositories())
if (!require("devtools")) install.packages("devtools", repos="http://cran.us.r-project.org")
if (!require("shiny")) install.packages("shiny", repos="http://cran.us.r-project.org")
if (!require("shinythemes")) install.packages("shinythemes", repos="http://cran.us.r-project.org")
if (!require("tidyverse")) install.packages("tidyverse", repos="http://cran.us.r-project.org")
if (!require("stringr")) install.packages("stringr", repos="http://cran.us.r-project.org")
if (!require("servr")) install.packages("servr", repos="http://cran.us.r-project.org")
if (!require("ggplot2")) install.packages("ggplot2", repos="http://cran.us.r-project.org")
if (!require("pheatmap")) install.packages("pheatmap", repos="http://cran.us.r-project.org")
if (!require("servr")) install.packages("servr", repos="http://cran.us.r-project.org")
if (!require("heatmaply")) install.packages("heatmaply", repos="http://cran.us.r-project.org")
if (!require("plotly")) install.packages("plotly", repos="http://cran.us.r-project.org")
if (!require("shinyjs")) install.packages("shinyjs", repos="http://cran.us.r-project.org")
if (!require("shinyBS")) install.packages("shinyBS", repos="http://cran.us.r-project.org")
if (!require("msigdbr")) install.packages("msigdbr", repos="http://cran.us.r-project.org")
if (!require("fontawesome")) install.packages("fontawesome", repos="http://cran.us.r-project.org")
if (!require("shinycssloaders")) install.packages("shinycssloaders", repos="http://cran.us.r-project.org")
if (!require("DT")) install.packages("DT", repos="http://cran.us.r-project.org")
if (!require("e1071")) install.packages("e1071", repos="http://cran.us.r-project.org")
if (!require("dplyr")) install.packages("dplyr", repos="http://cran.us.r-project.org")
if (!require("pbmcapply")) install.packages("pbmcapply", repos="http://cran.us.r-project.org")
if (!require("reshape2")) install.packages("reshape2", repos="http://cran.us.r-project.org")
if (!require("shinydashboard")) install.packages("shinydashboard", repos="http://cran.us.r-project.org")
if (!require("shinyalert")) install.packages("shinyalert", repos="http://cran.us.r-project.org")
if (!require("shinyFiles")) install.packages("shinyFiles", repos="http://cran.us.r-project.org")
if (!require("igraph")) install.packages("igraph", repos="http://cran.us.r-project.org")
if (!require("cluster")) install.packages("cluster", repos="http://cran.us.r-project.org")
if (!require("RANN")) install.packages("RANN", repos="http://cran.us.r-project.org")
if (!require("shinyWidgets")) install.packages("shinyWidgets", repos="http://cran.us.r-project.org")
if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")
if (!require("scales")) install.packages("scales", repos="http://cran.us.r-project.org")
if (!require("dtwclust")) install.packages("dtwclust", repos="http://cran.us.r-project.org")
if (!require("anndata")) install.packages("anndata", repos="http://cran.us.r-project.org")
if (!require("reticulate")) install.packages("reticulate", repos="http://cran.us.r-project.org")
if (!require("tibble")) install.packages("tibble", repos="http://cran.us.r-project.org")
if (!require("clustree")) install.packages("clustree", repos="http://cran.us.r-project.org")
if (!require("magrittr")) install.packages("magrittr", repos="http://cran.us.r-project.org")
if (!require("cowplot")) install.packages("cowplot", repos="http://cran.us.r-project.org")
if (!require("rhandsontable")) install.packages("rhandsontable", repos="http://cran.us.r-project.org")
if (!require("sctransform")) install.packages("sctransform", repos="http://cran.us.r-project.org")
if (!require("data.table")) install.packages("data.table", repos="http://cran.us.r-project.org")
if (!require("purrr")) install.packages("purrr", repos="http://cran.us.r-project.org")
if (!require("scater")) BiocManager::install("scater", update = FALSE)
if (!require("fgsea")) BiocManager::install("fgsea", update = FALSE)
if (!require("MOFA2")) BiocManager::install("MOFA2", update = FALSE)
if (!require("EnhancedVolcano")) BiocManager::install("EnhancedVolcano", update = FALSE)
if (!require("liana")) remotes::install_github('saezlab/liana')
if (!require("rGREAT")) install_github("jokergoo/rGREAT")
if (!require("biovizBase")) BiocManager::install("biovizBase", update = FALSE)
if (!require("GenomeInfoDb")) BiocManager::install("GenomeInfoDb", update = FALSE)
if (!require("BSgenome.Hsapiens.UCSC.hg19")) BiocManager::install("BSgenome.Hsapiens.UCSC.hg19", update = FALSE)
#if (!require("cytomapper")) BiocManager::install("cytomapper", update = FALSE)
#if (!require("FlowSOM")) BiocManager::install("FlowSOM", update = FALSE)
#if (!require("cytofkit2")) devtools::install_github("JinmiaoChenLab/cytofkit2", dependencies=TRUE)
if (!require("lisi")) devtools::install_github("immunogenomics/lisi", dependencies=TRUE)
if (!require("ktplots")) devtools::install_github('zktuong/ktplots', dependencies = TRUE)
if (!require("EnsDb.Hsapiens.v86")) BiocManager::install("EnsDb.Hsapiens.v86", update = FALSE)
#if (!require("flowCore")) BiocManager::install("flowCore", update = FALSE)
if (!require("harmony")) BiocManager::install("harmony")
#if (!require("rliger")) install.packages('rliger')
if (!require("cowplot")) install.packages('cowplot')
#if (!require("Rphenograph")) devtools::install_github("JinmiaoChenLab/Rphenograph", upgrade=TRUE)
#if (!require("CellChat")) devtools::install_github("sqjin/CellChat", upgrade=TRUE)
if (!require("sceasy")) devtools::install_github("cellgeni/sceasy", upgrade=TRUE)
if (!require("leidenbase")) devtools::install_github('cole-trapnell-lab/leidenbase')
#if (!require("monocle3")) devtools::install_github('cole-trapnell-lab/monocle3')
#if (!require("SPOTlight")) devtools::install_github("https://github.com/MarcElosua/SPOTlight", upgrade=TRUE)
if (!require("FastIntegration")) devtools::install_github("git@github.com:JinmiaoChenLab/FastIntegrate.git", upgrade=TRUE)
if (!require("hdf5r")) install.packages("hdf5r", repos="http://cran.us.r-project.org")
if (!require("Seurat")) install.packages("Seurat", repos="http://cran.us.r-project.org")
if (!require("bslib")) install.packages("bslib", repos="http://cran.us.r-project.org")
if (!require("shinyWidgets")) install.packages("shinyWidgets")
if (!require("SeuratWrappers")) remotes::install_github('satijalab/seurat-wrappers')
if (!require("Signac")) install.packages("Signac", repos="http://cran.us.r-project.org")
Raw
server.R
source('ui.R')
#source('cellphonedb.R')
#source('plot_heatmaps.R')
## max data size
#options(shiny.maxRequestSize = 1024^10)
options(shiny.maxRequestSize = 4096*4096*100*100)
options(shiny.launch.browser = T)
options(bitmapType = 'cairo')
options(timeout=1000000)
options(show.error.messages = F)
shinyServer(function(input, output, session) {
#use_python('/home/ezsinglecell/miniconda3/bin/python', required = NULL)
v <- reactiveValues(scData = NULL,
idents = NULL,
isPCAdone = NULL,
isUMAPdone = NULL,
isTSNEdone = NULL,
isTrajectorydone = NULL,
isCELLiDdone = NULL,
isCELLdone = NULL,
isGSEAdone = NULL,
pcGenes = NULL,
plotlySelection = NULL,
ips.markers = NULL,
markers = NULL,
selected_markers = NULL)
#celltypes <- NULL
prePlot <- function(){
while(names(dev.cur()) != "null device"){
dev.off()
}
}
observe({
#s <- event_data("plotly_selected")
#cells <- s[["key"]]
v$plotlySelection <- event_data("plotly_selected")[["key"]]
})
output$demo_image <- renderImage({
list(src = "www/Picture1.png",
width = 500,
height = 400)
}, deleteFile = FALSE)
output$demo_image1 <- renderImage({
list(src = "www/Picture2.png",
width = 330,
height = 400)
}, deleteFile = FALSE)
output$demo_image2 <- renderImage({
list(src = "www/overview1.png",
width = 500,
height = 600)
}, deleteFile = FALSE)
output$scrna_image1 <- renderImage({
list(src = "www/scRNA-seq.png",
height = 500)
}, deleteFile = FALSE)
#output$scrna_image2 <- renderImage({
# list(src = "www/scRNA-seq_2.png",
# height = 500)
#}, deleteFile = FALSE)
##-------------------scRNA-seq module-------------------
observeEvent(input$loadexample_tpm, {
withProgress(message="Loading example data...", value=0.5, {
tpmFiles_demo <- read.table("scRNA/pbmc.txt", header = T, row.names = 1, check.names = F)
sObj <- CreateSeuratObject(tpmFiles_demo,
project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
})
if (is.null(tpmFiles_demo)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
print(tpmFiles_demo)
#print(tpmFiles$name)
print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_tpm", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData <- tpmFiles_demo
v$scData1 <- sObj
})
}
})
observeEvent(input$loadexample_scH5, {
withProgress(message="Loading example data...", value=0.5, {
tpmFiles_demo <- Read10X_h5("scRNA/filtered_feature_bc_matrix_human.h5")
sObj <- CreateSeuratObject(tpmFiles_demo,
project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
})
if (is.null(tpmFiles_demo)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_scH5", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData <- tpmFiles_demo
v$scData1 <- sObj
})
}
})
observeEvent(input$loadexample_rds, {
withProgress(message="Loading example data...", value=0.5, {
tpmFiles_demo <- readRDS("scRNA/pbmc.rds")
sObj <- CreateSeuratObject(tpmFiles_demo@assays$RNA@counts,
project = input$projName)
print(sObj)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
})
if (is.null(tpmFiles_demo)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_rds", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData <- tpmFiles_demo@assays$RNA@counts
v$scData1 <- sObj
})
}
})
observeEvent(input$loadexample_h5ad, {
withProgress(message="Loading example data...", value=0.5, {
tpmFiles_demo <- read_h5ad("scRNA/pbmc.h5ad")
tpmFiles_demo = t(tpmFiles_demo$to_df())
sObj <- CreateSeuratObject(tpmFiles_demo, project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
print(sObj)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
print(sObj@meta.data)
})
if (is.null(tpmFiles_demo)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_h5ad", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData <- tpmFiles_demo
v$scData1 <- sObj
})
}
})
observeEvent(input$loadButton, {
if(input$scInput == "Raw Counts Matrix"){
tpmFiles <- input$tpmFiles
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else
{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFiles$datapath)
print(tpmFiles$name)
print(file.exists(paste(tpmFiles$datapath[1], "/", tpmFiles$name[1], sep="")))
exp.data <- read.table(tpmFiles$datapath,
sep="\t", header=TRUE, row.names=1, stringsAsFactors = F, check.names = F)
sObj <- CreateSeuratObject(exp.data,
project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
v$scData <- exp.data
v$scData1 <- sObj
label1 <- "Data loaded"
updateActionButton(inputId = "loadButton", label = label1)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput == "10X cellranger"){
scH5 <- input$scH5
if (is.null(scH5)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(scH5$datapath)
print(scH5$name)
print(file.exists(paste(scH5$datapath[1], "/", scH5$name[1], sep="")))
exp.data <- Read10X_h5(scH5$datapath)
sObj <- CreateSeuratObject(exp.data,
project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
v$scData <- exp.data
v$scData1 <- sObj
label1 <- "Data loaded"
updateActionButton(inputId = "loadButton", label = label1)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput == "rds object"){
rds <- input$rds
if (is.null(rds)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(rds$datapath)
print(rds$name)
print(file.exists(paste(rds$datapath[1], "/", rds$name[1], sep="")))
exp.data <- readRDS(rds$datapath)
sObj <- CreateSeuratObject(exp.data@assays$RNA@counts,
project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
v$scData <- exp.data@assays$RNA@counts
v$scData1 <- sObj
label1 <- "Data loaded"
updateActionButton(inputId = "loadButton", label = label1)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput == "h5ad"){
h5ad <- input$h5ad
if (is.null(h5ad)){
v$scData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(h5ad$datapath)
print(h5ad$name)
print(file.exists(paste(h5ad$datapath[1], "/", h5ad$name[1], sep="")))
exp.data <- read_h5ad(h5ad$datapath)
exp.data = t(exp.data$to_df())
sObj <- CreateSeuratObject(exp.data, project = input$projName)
#min.genes = input$min.genes,
#min.cells = input$min.cells)
sObj[["percent.mt"]] <- PercentageFeatureSet(sObj, pattern = "^MT-")
v$scData <- exp.data
v$scData1 <- sObj
label1 <- "Data loaded"
updateActionButton(inputId = "loadButton", label = label1)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
})
output$downloadCount <- downloadHandler(
filename = function(){"counts_table.csv"},
content = function(fname){
withProgress(message="Downloading counts data...", value=0, {
write.csv(v$scData, fname)
})
}
)
observeEvent(input$filter_seurat, {
if(input$scInput == "Raw Counts Matrix"){
tpmFiles <- input$tpmFiles
tpmFiles <- v$scData
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scData)
v$scData1 <- subset(v$scData1, subset = nFeature_RNA > input$ob1 & nFeature_RNA < input$ob2 & percent.mt < input$ob3)
print(v$scData1)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput == "10X cellranger"){
scH5 <- input$scH5
scH5 <- v$scData
if (is.null(scH5)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scData)
v$scData1 <- subset(v$scData1, subset = nFeature_RNA > input$ob1 & nFeature_RNA < input$ob2 & percent.mt < input$ob3)
print(v$scData1)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput == "rds object"){
rds <- input$rds
rds <- v$scData
if (is.null(rds)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scData)
v$scData1 <- subset(v$scData1, subset = nFeature_RNA > input$ob1 & nFeature_RNA < input$ob2 & percent.mt < input$ob3)
print(v$scData1)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput == "h5ad"){
h5ad <- input$h5ad
h5ad <- v$scData
if (is.null(h5ad)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scData)
v$scData1 <- subset(v$scData1, subset = nFeature_RNA > input$ob1 & nFeature_RNA < input$ob2 & percent.mt < input$ob3)
print(v$scData1)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
})
observeEvent(input$reset_scRNA, {
session$reload()
print("Reset done")
})
#})
output$logo <- renderImage({
return(list(
src = "inst/extdata/logo.png",
contentType = "image/png",
alt = "Singapore Immunology Network"
))
}, deleteFile = FALSE)
opendir <- function(dir = getwd()){
if (.Platform['OS.type'] == "windows"){
shell.exec(dir)
} else {
system(paste(Sys.getenv("R_BROWSER"), dir))
}
}
##---------------QC of scRNA-seq-------------------
output$countdataDT <- renderDataTable({
if(!is.null(v$scData))
{
if(ncol(v$scData) > 20 )
as.matrix(v$scData[,1:20])
}
}, server = FALSE)
plotInput <- reactive({
p <- VlnPlot(v$scData1, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3) + NoLegend()
})
output$nFeature_RNAPlot <- renderPlot({
if(is.null(v$scData1)){
plotly_empty()
}else{
print(plotInput())
}
})
output$download_nFeature_RNA <- downloadHandler(
filename = function(){"QC violin plots.png"},
content = function(fname){
ggsave(fname,plotInput(), width = 15, height = 10)
}
)
## FeatureScatter plot
plotInput3 <- reactive({
p <- FeatureScatter(v$scData1, "nCount_RNA", "nFeature_RNA")
})
output$FeatureScatterPlot1 <- renderPlotly({
if(is.null(v$scData1)){
plotly_empty()
}else{
print(plotInput3())
}
})
output$download_FeatureScatterPlot1 <- downloadHandler(
filename = function(){"FeatureScatterPlot1.png"},
content = function(fname){
ggsave(fname,plotInput3(), height = 7, width = 7)
}
)
plotInput4 <- reactive({
p <- FeatureScatter(v$scData1, "nCount_RNA", "percent.mt")
})
output$FeatureScatterPlot2 <- renderPlotly({
if(is.null(v$scData1)){
plotly_empty()
}else{
print(plotInput4())
}
})
output$download_FeatureScatterPlot2 <- downloadHandler(
filename = function(){"FeatureScatterPlot2.png"},
content = function(fname){
ggsave(fname,plotInput4(), height = 7, width = 7)
}
)
observeEvent(input$findVarGenes, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding variable genes...", value = 0, {
if(input$norm1 == "LogNormalize"){
v$scData1 <- NormalizeData(v$scData1)
v$scData1 <- FindVariableFeatures(v$scData1,
mean.function = ExpMean,
dispersion.function = LogVMR,
nfeatures = input$var.genes,
selection.method = input$selection.method)
all.genes <- rownames(v$scData1)
v$scData1 <- ScaleData(v$scData1, features = all.genes)
incProgress(0.5)
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData1@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scData1)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scData1), 10)
variable_feature1 <- VariableFeaturePlot(v$scData1)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("Highly variable features identified", "Highly variable features identified, please perform PCA", type = "success", imageWidth = 10, imageHeight = 10)
#dev.off()
})
}
}
output$VarGenes <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scData1)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Seurat_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scData1)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scData1@assays$RNA@var.features, file = txtfile)
})
}
})
}
else if(input$norm1 == "SCTransform"){
v$scData1 <- SCTransform(v$scData1, variable.features.n = input$var.genes, vars.to.regress = "percent.mt", verbose = FALSE, conserve.memory = T)
incProgress(0.5)
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scData1)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scData1), 10)
variable_feature1 <- VariableFeaturePlot(v$scData1)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("SCTransform done", "SCTransform done, please perform PCA", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
output$VarGenes <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scData1)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Seurat_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scData1)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scData1@assays$RNA@var.features, file = txtfile)
})
}
})
}
})
}
})
output$name <- renderPrint({
s <- event_data("plotly_selected")
c(s[["key"]], class(s[["key"]]))
})
##---------------PCA of scRNA-seq-------------------
# PCA plot
observeEvent(input$doPCA, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Scaling Data...", value = 0,{
incProgress(0.5, message = "Running PCA...")
#if(input$assays1 == "LogNormalization"){
if(length(colnames(v$scData1@assays$RNA))<5000){
v$scData1 <- RunPCA(v$scData1, features = rownames(v$scData1))
}
else if(length(colnames(v$scData1@assays$RNA))>=5000){
v$cells <- sketchData(v$scData1, percent = 0.25, idents = NULL, do.PCA = TRUE, dimPC = 30)
v$scData1 <- v$scData1[,v$cells]
v$scData1 <- RunPCA(v$scData1, features = rownames(v$scData1))
}
#print(v$scData1[["pca"]], dims = 1:5, nfeatures = 5)
#else if(input$assays1 == "SCTransform"){
# v$scData1 <- RunPCA(v$scData1, features = rownames(v$scData1), assay = "SCT")
# print(v$scData1[["pca"]], dims = 1:5, nfeatures = 5)
#}
v$isPCAdone <- TRUE
PCA_plot <- DimPlot(v$scData1, reduction = "pca", label = T)
print(PCA_plot)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scData1)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
shinyalert("PCA performed", "PCA performed, please perform clustering", type = "success", imageWidth = 10, imageHeight = 10)
label1 <- "PCA Done"
updateActionButton(inputId = "doPCA", label = label1)
})
}
})
plotPCA <- reactive({
if(is.null(v$scData1) || is.null(v$isPCAdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "pca", label = T)
})
}
})
output$PCA2DPlot <- renderPlotly({
plotPCA()
})
output$download_PCA <- downloadHandler(
filename = function(){"PCAPlot.png"},
content = function(fname){
ggsave(fname,PCAPlot(v$scData1), height = 7, width = 7)
}
)
output$download_PCA_embedding <- downloadHandler(
filename = function(){"pca_embeddings.csv"},
content = function(fname){
withProgress(message="Downloading counts data...", value=0, {
write.csv(v$scData1@reductions$pca@cell.embeddings, fname)
})
}
)
# Viz plot
plotViz <- reactive({
if(is.null(v$scData1) || is.null(v$isPCAdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- VizDimLoadings(v$scData1, dims = as.numeric(input$select.pc))
})
}
})
output$vizPlot <- renderPlot({
print(plotViz())
})
output$download_vizPlot <- downloadHandler(
filename = function(){"VizLoading_Plot.png"},
content = function(fname){
ggsave(fname,plotViz(), height = 7, width = 7)
}
)
plotPCHeatmap <- reactive({
if(is.null(v$scData1) || is.null(v$isPCAdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimHeatmap(v$scData1, dims = as.numeric(input$select.pc))
})
}
})
output$PCHeatmap <- renderPlot({
print(plotPCHeatmap())
})
output$download_PCHeatmap <- downloadHandler(
filename = function(){"PC Heatmap.png"},
content = function(fname){
png(fname)
DimHeatmap(v$scData1, dims = as.numeric(input$select.pc))
dev.off()
}
)
output$PCtable <- DT::renderDataTable({
if(is.null(v$scData1) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$download_PCTable <- downloadHandler(
filename = function(){"PC_table.csv"},
content = function(fname){
withProgress(message="Downloading PC Table...", value=0, {
write.csv(v$pcGenes, fname)
})
}
)
plotElbow <- reactive({
if(is.null(v$scData1) || is.null(v$isPCAdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- ElbowPlot(v$scData1, ndims = 50)
})
}
})
output$Elbow <- renderPlot({
print(plotElbow())
})
output$download_Elbow <- downloadHandler(
filename = function(){"Elbow plot.png"},
content = function(fname){
ggsave(fname,plotElbow(), height = 7, width = 7)
}
)
##---------------Clustering of scRNA-seq-------------------
observeEvent(input$findCluster, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
#if(input$assays1 == "LogNormalization"){
#DefaultAssay(v$scData1) <- "RNA"
v$scData1 <- FindNeighbors(v$scData1, dims = 1:input$dim.used, nn.method = "rann")
v$scData1 <- FindClusters(v$scData1, resolution = input$clus.res)
print(v$scData1@meta.data)
output$cluster.done <- renderText(paste0("Clustering done!"))
#}
#else if(input$assays1 == "SCTransform"){
# DefaultAssay(v$scData1) <- "SCT"
# v$scData1 <- FindNeighbors(v$scData1, dims = 1:input$dim.used, assay = "SCT", nn.method = "rann")
# v$scData1 <- FindClusters(v$scData1, resolution = input$clus.res)
# output$cluster.done <- renderText(paste0("Clustering done!"))
#}
v$isClusterdone <- TRUE
shinyalert("Clustering performed", "Clustering performed, please identify optimum number of clusters", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
plotCluster <- reactive({
p <- DimPlot(v$scData1, reduction = "umap", label = T) + NoLegend()
})
output$Cluster2DPlot_1 <- renderPlotly({
if(is.null(v$scData1) || is.null(v$isClusterdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
print(plotCluster())
})
}
})
output$download_Cluster <- downloadHandler(
filename = function(){"Cluster plot.png"},
content = function(fname){
ggsave(fname,plotCluster(), height = 7, width = 7)
}
)
output$download_ClusterTable <- downloadHandler(
filename = function(){"Cluster_table.csv"},
content = function(fname){
withProgress(message="Downloading Cluster Table...", value=0, {
write.csv(v$scData1$seurat_clusters, fname)
})
}
)
##---------Find Optimum Cluster Resolution---------##
observeEvent(input$findoptimumCluster, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding optimum resolution...", value = 0.3, {
#if(input$assays1 == "LogNormalization"){
#DefaultAssay(v$scData1) <- "RNA"
reses<-seq(from = input$clus.res_a, to = input$clus.res_b, by = 0.1)
for (res in reses){
v$scData1<-FindClusters(v$scData1, resolution = res)
nores<-gsub(pattern = ".", replacement = "", res, fixed = T)
}
print(v$scData1@meta.data)
output$optimumcluster.done <- renderText(paste0("Clustering done!"))
#}
#else if(input$assays1 == "SCTransform"){
# DefaultAssay(v$scData1) <- "SCT"
# v$scData1 <- FindNeighbors(v$scData1, dims = 1:input$dim.used, assay = "SCT", nn.method = "rann")
# v$scData1 <- FindClusters(v$scData1, resolution = input$clus.res)
# output$cluster.done <- renderText(paste0("Clustering done!"))
#}
v$isOptimumClusterdone <- TRUE
shinyalert("Number of clusters identified at different resolution", "Number of clusters identified at different resolution, please perform UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
plotOptimumCluster <- reactive({
if(is.null(v$scData1) || is.null(v$isOptimumClusterdone)){
plotly_empty()
}else{
withProgress(message="Plotting Clustering tree...", value=0, {
p <- clustree(v$scData1)
})
}
})
output$OptimumCluster2DPlot_1 <- renderPlot({
print(plotOptimumCluster())
})
output$download_OptimumCluster <- downloadHandler(
filename = function(){"Clustree plot.png"},
content = function(fname){
ggsave(fname,plotOptimumCluster(), height = 7, width = 7)
}
)
output$download_OptimumClusterTable <- downloadHandler(
filename = function(){"Clustering_different_resolution.csv"},
content = function(fname){
withProgress(message="Downloading Cluster Table at different resolution...", value=0, {
write.csv(v$scData1@meta.data, fname)
})
}
)
##---------------UMAP of scRNA-seq-------------------
observeEvent(input$doUmap, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0.3, {
#if(input$assays1 == "LogNormalization"){
v$scData1 <- RunUMAP(v$scData1, dims = 1:input$dim.used, spread = 1)
output$Umap.done <- renderText(paste0("UMAP done!"))
#}
#else if(input$assays1 == "SCTransform"){
# v$scData1 <- RunUMAP(v$scData1, dims = 1:input$dim.used, assay = "SCT", spread = 1)
# output$Umap.done <- renderText(paste0("UMAP done!"))
#}
v$isUMAPdone <- TRUE
shinyalert("UMAP performed", "UMAP performed, please perform tSNE", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
plotUMAP <- reactive({
if(is.null(v$scData1) || is.null(v$isUMAPdone)){
return(NULL)
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "umap", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_2d_plot_1 <- renderPlotly({
plotUMAP()
})
output$download_UMAP <- downloadHandler(
filename = function(){"UMAP plot.png"},
content = function(fname){
ggsave(fname,plotUMAP(), height = 7, width = 7)
}
)
output$download_UMAP_embedding <- downloadHandler(
filename = function(){"UMAP Embeddings.csv"},
content = function(fname){
withProgress(message="Downloading UMAP Embeddings...", value=0, {
write.csv(v$scData1@reductions$umap@cell.embeddings, fname)
})
}
)
##---------------TSNE of scRNA-seq-------------------
output$perplex.option <- renderUI({
if(is.null(v$isPCAdone)){
return(NULL)
}else{
##perplexity test
n.cells <- isolate(nrow(v$scData1@reductions$pca@cell.embeddings))
max.perplex <- as.integer((n.cells - 1)/3)
numericInput("perplexity",
label = "Perplexity",
value = if(max.perplex <30) max.perplex else 30,
min = 0,
max = max.perplex)
}
})
observeEvent(input$doTsne, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running tSNE...", value = 0.3, {
#if(input$assays1 == "LogNormalization"){
v$scData1 <- RunTSNE(v$scData1, dims = 1:input$dim.used, perplexity = input$perplexity)
output$Tsne.done <- renderText(paste0("TSNE done!"))
#}
#else if(input$assays1 == "SCTransform"){
# v$scData1 <- RunTSNE(v$scData1, dims = 1:input$dim.used, perplexity = input$perplexity, assay = "SCT")
# output$Tsne.done <- renderText(paste0("TSNE done!"))
#}
v$isTSNEdone <- TRUE
shinyalert("tSNE performed", "tSNE performed, please perform celltype annotation", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
plotTsne <- reactive({
if(is.null(v$scData1) || is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "tsne", label = T, label.size = 3) + NoLegend()
})
}
})
output$Tsne_2d_plot_1 <- renderPlotly({
plotTsne()
})
output$download_Tsne <- downloadHandler(
filename = function(){"tSNE plot.png"},
content = function(fname){
ggsave(fname,plotTsne(), height = 7, width = 7)
}
)
output$download_Tsne_embedding <- downloadHandler(
filename = function(){"tSNE Embeddings.csv"},
content = function(fname){
withProgress(message="Downloading tSNE Embeddings...", value=0, {
write.csv(v$scData1@reductions$tsne@cell.embeddings, fname)
})
}
)
##---------------Run CELLiD of scRNA-seq-------------------
observeEvent(input$doCELLiD, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running CELLiD...", value = 0.3, {
ref = readRDS('ref.rds')
v$scData1.rna.data.average = AverageExpression(v$scData1)
v$scData1.rna.data.average = round(v$scData1.rna.data.average$RNA, 2)
#v$scData1.rna.data.average = data.frame(v$scData1.rna.data.average$RNA)
if(input$cellatlas == "all"){
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, ref)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, adipose1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, adrenal_gland1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, blood1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, bone_marrow1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, brain1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, breast1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, breast_milk1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, eye1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, gut1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, heart1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, kidney1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, liver1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, lung1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, pancreas1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, PDAC1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, skin1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, testis1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, thymus1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
v$res = FastIntegration::CELLiD(v$scData1.rna.data.average, tonsil1)
print(v$res)
v$scData1$primary.predict = v$res[as.numeric(v$scData1$seurat_clusters),1]
v$scData1$secondary.predict = v$res[as.numeric(v$scData1$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scData1@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
v$isCELLiDdone <- TRUE
shinyalert("Celltype annotation performed", "Celltype annotation performed, please perform cell-cell similarity", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_cellid <- renderPlotly({
if(is.null(v$scData1) || is.null(v$isCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scData1, reduction = "umap", group.by = "primary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_cellid1 <- renderPlotly({
if(is.null(v$scData1) || is.null(v$isCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scData1, reduction = "umap", group.by = "secondary.predict", label = T, label.size = 3)
})
}
})
output$ct.table <- DT::renderDataTable(
v$res, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
#onclick("commitButton",{
# proxy=dataTableProxy("ct.table")
# replaceData(ct.table,ds)
#})
plotCELLiD <- reactive({
if(is.null(v$scData1) || is.null(v$isCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "umap", group.by = "primary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_cellid <- renderPlotly({
plotCELLiD()
})
output$download_Umap_cellid <- downloadHandler(
filename = function(){"Celltype identification plot (Primary prediction).png"},
content = function(fname){
ggsave(fname,plotCELLiD(), height = 7, width = 7)
}
)
plotCELLiD1 <- reactive({
if(is.null(v$scData1) || is.null(v$isCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "umap", group.by = "secondary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_cellid1 <- renderPlotly({
plotCELLiD1()
})
output$download_Umap_cellid1 <- downloadHandler(
filename = function(){"Celltype identification plot (Secondary prediction).png"},
content = function(fname){
ggsave(fname,plotCELLiD1(), height = 7, width = 7)
}
)
output$download_cellid_prediction <- downloadHandler(
filename = function(){"CELLiD_predictions.csv"},
content = function(fname){
withProgress(message="Downloading CELLiD predictions...", value=0, {
write.csv(v$res, fname)
})
}
)
recodeValues <- reactive({
if(is.null(v$scData1) || is.null(v$isCELLiDdone)) {return()}
uniqueVars <- v$res[,1]
print(uniqueVars)
convert <- cbind(uniqueVars, uniqueVars)
print(uniqueVars)
print(convert)
colnames(convert) <- c("Original ID", "newID")
convert
})
output$recoding <- DT::renderDataTable({
recodeValues()
}, server = FALSE)
plotCELLiD2 <- reactive({
if(is.null(v$scData1) || is.null(v$isRenameCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "umap", group.by = "newID", label = T, label.size = 3) + NoLegend()
p
})
}
})
output$Umap_cellid2 <- renderPlotly({
plotCELLiD2()
})
fname = 'CELLiD_predictions.csv'
values = list()
setHot = function(x) values[["hot"]] <<- x
observeEvent(input$commitButton, {
if (!is.null(values[["hot"]])) {
write.csv(values[["hot"]], fname)
print(fname)
##### use the recoded values to rename values throughout the entire data frame
recodes <- read.csv(fname, stringsAsFactors = FALSE)
for(i in 1:nrow(recodes)){
if(recodes[i,2] != recodes[i,3]){
v$res[v$res == recodes[i,2]] <- recodes[i,3]
}
}
test <- read.csv('CELLiD_predictions.csv', header = T, row.names = 1)
as.numeric(rownames(test)) -> test$seurat_clusters
test[3] <- test[3]-1
test <- test[,-1]
print(test)
test$seurat_clusters <- as.factor(test$seurat_clusters)
meta <- v$scData1@meta.data
rownames(meta) -> meta$cellID
new <- merge(x=meta,y=test,by="seurat_clusters")
colnames(new) <- gsub('.x','',names(new))
colnames(new) <- gsub('.y','',names(new))
new$cellID -> rownames(new)
print(new)
v$scData1 <- AddMetaData(v$scData1, metadata = new)
print(v$scData1@meta.data)
v$isRenameCELLiDdone <- TRUE
#output$newVars <- DT::renderDataTable({
# v$res
#})
}
})
output$hot = renderRHandsontable({
if (!is.null(input$hot)) {
DF = hot_to_r(input$hot)
} else {
if(is.null(v$res)) {return()}
DF = recodeValues()
}
setHot(DF)
rhandsontable(DF) %>%
hot_table(highlightCol = TRUE, highlightRow = TRUE)
})
output$ct.gene.select <- renderUI({
if(is.null(v$scData1)|| is.null(v$isCELLiDdone)){
plotly_empty()
}else{
selectInput("ct.gene", label = "Gene to visualise",
choices = rownames(v$scData1@assays$RNA@counts))
}
})
observeEvent(input$Vis_seurat1, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData1 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Visualizing...", value=0, {
v$isVisCELLiDdone <- TRUE
})
}
})
plotViolin1 <- reactive({
if(is.null(v$scData1)|| is.null(v$isVisCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating Violin Plot...", value=0, {
print(v$scData1)
v$scData1$primary.predict -> Idents(v$scData1)
VlnPlot(v$scData1, input$ct.gene)
})
}
})
output$ct.gene.plot <- renderPlotly({
plotViolin1()
})
output$download_violn1 <- downloadHandler(
filename = function(){"Violin plot (CELLiD).png"},
content = function(fname){
ggsave(fname,plotViolin(), height = 7, width = 7)
}
)
plotFeature1 <- reactive({
if(is.null(v$scData1)|| is.null(v$isVisCELLiDdone)){
plotly_empty()
}else{
withProgress(message="Generating Violin Plot...", value=0, {
print(v$scData1)
v$scData1$primary.predict -> Idents(v$scData1)
FeaturePlot(v$scData1, input$ct.gene)
})
}
})
output$ct.gene1.plot <- renderPlotly({
plotFeature1()
})
output$download_feature1 <- downloadHandler(
filename = function(){"Feature plot (CELLiD).png"},
content = function(fname){
ggsave(fname,plotFeature1(), height = 7, width = 7)
}
)
#onclick("BRefresh",{
# proxy=dataTableProxy("OPreview")
# replaceData(proxy,ds)
#})
##---------------Run celltypist-----------------##
observeEvent(input$doCelltypist, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Celltypist...", value = 0.3, {
sc <- reticulate::import("scanpy", convert = FALSE)
ct <- reticulate::import("celltypist", convert = FALSE)
sceasy::convertFormat(v$scData1, from = "seurat", to = "anndata", outFile = 'ct_scrna.h5ad')
v$adata = sc$read_h5ad('ct_scrna.h5ad')
v$res = ct$annotate(filename = 'ct_scrna.h5ad', model = input$celltypistatlas, majority_voting=T)
print(v$res)
v$adata = v$res$to_adata()
print("fff")
v$adata$obs$to_csv('celltypist_predict.csv')
v$meta <- read.csv('celltypist_predict.csv', header = T, row.names = 1)
v$scData1 <- AddMetaData(v$scData1, metadata = v$meta)
v$scData1$primary.predict <- v$scData1$majority_voting
v$scData1$secondary.predict <- v$scData1$predicted_labels
v$res1 <- v$scData1@meta.data
print(v$scData1@meta.data)
v$isCelltypistdone <- TRUE
shinyalert("Celltype annotation performed", "Celltype annotation performed, please perform cell-cell similarity", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_celltypist <- renderPlotly({
if(is.null(v$scData1) || is.null(v$isCelltypistdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scData1, reduction = "umap", group.by = "majority_voting", label = T, label.size = 3)
})
}
})
output$Umap_celltypist1 <- renderPlotly({
if(is.null(v$scData1) || is.null(v$isCelltypistdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scData1, reduction = "umap", group.by = "predicted_labels", label = T, label.size = 3)
})
}
})
output$celltypist.table <- DT::renderDataTable(
v$meta, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
plotCELLTypist <- reactive({
if(is.null(v$scData1) || is.null(v$isCelltypistdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "umap", group.by = "majority_voting", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_celltypist <- renderPlotly({
plotCELLTypist()
})
output$download_Umap_celltypist <- downloadHandler(
filename = function(){"Celltypist plot (Majority voting).png"},
content = function(fname){
ggsave(fname,plotCELLTypist(), height = 7, width = 7)
}
)
plotCELLTypist1 <- reactive({
if(is.null(v$scData1) || is.null(v$isCelltypistdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
p <- DimPlot(v$scData1, reduction = "umap", group.by = "predicted_labels", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_celltypist1 <- renderPlotly({
plotCELLTypist1()
})
output$download_Umap_celltypist1 <- downloadHandler(
filename = function(){"Celltypist plot (Predicted labels).png"},
content = function(fname){
ggsave(fname,plotCELLTypist1(), height = 7, width = 7)
}
)
output$download_celltypist_prediction <- downloadHandler(
filename = function(){"Celltypist predictions.csv"},
content = function(fname){
withProgress(message="Downloading Celltypist predictions...", value=0, {
write.csv(v$meta, fname)
})
}
)
plotCELLTypist2 <- reactive({
if(is.null(v$scData1) || is.null(v$isRenameCELLTypistdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
test <- read.csv('CELLTypist predictions.csv', header = T, row.names = 1)
test <- test[c(10,8,5)]
#test[4] <- test[4]
#test <- test[,-1]
test$seurat_clusters <- as.factor(test$seurat_clusters)
meta <- v$scData1@meta.data
rownames(meta) -> meta$cellID
new <- merge(x=meta,y=test,by="seurat_clusters")
new$cellID -> rownames(new)
v$scData1 <- AddMetaData(v$scData1, metadata = new)
p <- DimPlot(v$scData1, reduction = "umap", group.by = "newID", label = T, label.size = 3) + NoLegend()
p
})
}
})
output$Umap_celltypist2 <- renderPlotly({
plotCELLTypist2()
})
output$celltypist.gene.select <- renderUI({
if(is.null(v$scData1)|| is.null(v$isCelltypistdone)){
return(NULL)
}else{
selectInput("celltypist.gene", label = "Gene to visualise",
choices = rownames(v$scData1@assays$RNA@counts))
}
})
output$celltypist.gene.plot <- renderPlotly({
if(is.null(v$scData1)|| is.null(v$isCelltypistdone)){
return(NULL)
}else{
withProgress(message="Generating Plot...", value=0, {
print(v$scData1)
v$scData1$majority_voting -> Idents(v$scData1)
VlnPlot(v$scData1, input$celltypist.gene)
})
}
})
output$celltypist.gene1.plot <- renderPlotly({
if(is.null(v$scData1)|| is.null(v$isCelltypistdone)){
return(NULL)
}else{
withProgress(message="Generating Plot...", value=0, {
FeaturePlot(v$scData1, input$celltypist.gene)
})
}
})
observe({
if(input$dodeconv_spatial > 0){
print('2')
session$sendCustomMessage("myCallbackHandler3", "2")
}
})
observe({
if(input$dodeconv_spatial1 > 0){
print('2')
session$sendCustomMessage("myCallbackHandler3a", "2")
}
})
observe({
if(input$dodeconv_spatial_intg > 0){
print('2')
session$sendCustomMessage("myCallbackHandler6", "2")
}
})
observe({
if(input$doct_atac_intg > 0){
print('2')
session$sendCustomMessage("myCallbackHandler7", "2")
}
})
observe({
if(input$doct_atac > 0){
print('2')
session$sendCustomMessage("myCallbackHandler5", "2")
}
})
observe({
if(input$doct_atac1 > 0){
print('2')
session$sendCustomMessage("myCallbackHandler5a", "2")
}
})
##---------------Cell-cell similarity of scRNA-seq-------------------
observeEvent(input$cell_cell, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating heatmap...", value=0, {
if(input$cell1 == "primary.predict"){
v$scData1$primary.predict -> Idents(v$scData1)
v$scData1.rna.data.average1 = AverageExpression(v$scData1)
v$scData1.rna.data.average1 = data.frame(v$scData1.rna.data.average1$RNA)
print(v$scData1.rna.data.average1)
v$cor <- cor(v$scData1.rna.data.average1, method = input$corr_method)
print(v$cor)
output$CELL.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLdone <- TRUE
}
if(input$cell1 == "seurat_clusters"){
v$scData1$seurat_clusters -> Idents(v$scData1)
v$scData1.rna.data.average1 = AverageExpression(v$scData1)
v$scData1.rna.data.average1 = data.frame(v$scData1.rna.data.average1$RNA)
print(v$scData1.rna.data.average1)
v$cor <- cor(v$scData1.rna.data.average1, method = input$corr_method)
rownames(v$cor) <- substr(rownames(v$cor),2,nchar(rownames(v$cor)))
colnames(v$cor) <- substr(colnames(v$cor),2,nchar(colnames(v$cor)))
print(v$cor)
output$CELL.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLdone <- TRUE
}
if(input$cell1 == "newID"){
v$scData1$newID -> Idents(v$scData1)
v$scData1.rna.data.average1 = AverageExpression(v$scData1)
v$scData1.rna.data.average1 = data.frame(v$scData1.rna.data.average1$RNA)
print(v$scData1.rna.data.average1)
v$cor <- cor(v$scData1.rna.data.average1, method = input$corr_method)
print(v$cor)
output$CELL.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLdone <- TRUE
}
})
}
})
plotCELL <- reactive({
if(is.null(v$scData1.rna.data.average1) || is.null(v$isCELLdone)){
plotly_empty()
}else{
withProgress(message="Generating Celltype similarity plot...", value=0, {
p <- heatmaply(as.matrix(v$cor), cexRow = 0.8, cexCol = 0.8, margins = c(10,10), k_col =2, k_row = 2, colors = rev(RColorBrewer::brewer.pal(9, "RdBu")))
})
}
})
output$cell_cell_sim <- renderPlotly({
plotCELL()
})
output$download_cell_cell_sim <- downloadHandler(
filename = "Celltype similarity.png",
content = function(file) {
png(file)
heatmap(as.matrix(v$cor), col = RColorBrewer::brewer.pal(9, "RdBu"), cexRow = 0.8, cexCol = 0.8, margins = c(10,10))
dev.off()
}
)
output$cor.table <- DT::renderDataTable(
v$cor, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_cor.table <- downloadHandler(
filename = function(){"Celltype_similarity.csv"},
content = function(fname){
withProgress(message="Downloading celltype_similarity...", value=0, {
write.csv(v$cor, fname)
})
}
)
##---------------Sub-cluster analysis--------------##
output$subcluster.gene.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
selectInput("cluster_subcluster", label = "Cluster to subcluster",
choices = unique(v$scData1$seurat_clusters), selected = unique(v$scData1$seurat_clusters)[1])
}
})
observeEvent(input$subcluster, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Performing subcluster analysis...", value=0, {
v$scData1$seurat_clusters -> Idents(v$scData1)
v$scData1.subset <- FindSubCluster(v$scData1, input$cluster_subcluster, "RNA_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res, algorithm = 1)
print(v$scData1.subset)
output$Subcluster.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone <- TRUE
})
}
})
plotSubclster <- reactive({
if(is.null(v$scData1.subset) || is.null(v$isSubclusterdone)){
plotly_empty()
}else{
withProgress(message="Generating Celltype similarity plot...", value=0, {
p <- DimPlot(v$scData1.subset, reduction = "umap", group.by = "sub.cluster", label = T, label.size = 2.5)
})
}
})
output$subcluster_plot <- renderPlotly({
plotSubclster()
})
output$download_subcluster <- downloadHandler(
filename = function(){"Subcluster plot.png"},
content = function(fname){
ggsave(fname,plotSubclster(), height = 7, width = 7)
}
)
##---------------DEGs of scRNA-seq-------------------
observeEvent(input$doDeg, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
if(input$deg1 == "seurat_clusters"){
v$scData1$seurat_clusters -> Idents(v$scData1)
if (input$norm1 == "LogNormalize"){
ips.markers <- FindAllMarkers(v$scData1, only.pos = FALSE, min.pct = input$min_pct, logfc.threshold = input$logfc, assay = "RNA", test.use = input$test.use)
v$ips.markers <- ips.markers
v$isDEGdone <- TRUE
}
if (input$norm1 == "SCTransform"){
ips.markers <- FindAllMarkers(v$scData1, only.pos = FALSE, min.pct = input$min_pct, logfc.threshold = input$logfc, assay = "SCT", test.use = input$test.use)
v$ips.markers <- ips.markers
v$isDEGdone <- TRUE
}
shinyalert("DEGs estimated", "DEGs estimated, please perform do data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
if(input$deg1 == "primary.predict"){
v$scData1$primary.predict -> Idents(v$scData1)
if (input$norm1 == "LogNormalize"){
ips.markers <- FindAllMarkers(v$scData1, only.pos = FALSE, min.pct = input$min_pct, logfc.threshold = input$logfc, assay = "RNA", test.use = input$test.use)
v$ips.markers <- ips.markers
v$isDEGdone <- TRUE
}
if (input$norm1 == "SCTransform"){
ips.markers <- FindAllMarkers(v$scData1, only.pos = FALSE, min.pct = input$min_pct, logfc.threshold = input$logfc, assay = "SCT", test.use = input$test.use)
v$ips.markers <- ips.markers
v$isDEGdone <- TRUE
}
shinyalert("DEGs estimated", "DEGs estimated, please perform do data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
if(input$deg1 == "newID"){
v$scData1$newID -> Idents(v$scData1)
if (input$norm1 == "LogNormalize"){
ips.markers <- FindAllMarkers(v$scData1, only.pos = FALSE, min.pct = input$min_pct, logfc.threshold = input$logfc, assay = "RNA", test.use = input$test.use)
v$ips.markers <- ips.markers
v$isDEGdone <- TRUE
}
if (input$norm1 == "SCTransform"){
ips.markers <- FindAllMarkers(v$scData1, only.pos = FALSE, min.pct = input$min_pct, logfc.threshold = input$logfc, assay = "SCT", test.use = input$test.use)
v$ips.markers <- ips.markers
v$isDEGdone <- TRUE
}
shinyalert("DEGs estimated", "DEGs estimated, please perform do data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
observeEvent(input$Vis_seurat, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData1 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Visualizing...", value=0, {
v$isVisdone <- TRUE
})
}
})
output$deg.gene.select <- renderUI({
if(is.null(v$scData1)){
return(NULL)
}else{
selectInput("deg.gene", label = "Gene to visualise",
choices = rownames(v$scData1@assays$RNA@counts), selected = rownames(v$scData1@assays$RNA@counts)[50])
}
})
plotViolin <- reactive({
if(is.null(v$scData1) || is.null(v$isVisdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg2 == "seurat_clusters"){
print(v$scData1)
v$scData1$seurat_clusters -> Idents(v$scData1)
print(v$scData1)
VlnPlot(v$scData1, input$deg.gene)
}
else if(input$deg2 == "primary.predict"){
v$scData1$primary.predict -> Idents(v$scData1)
VlnPlot(v$scData1, input$deg.gene)
}
else if(input$deg2 == "newID"){
v$scData1$newID -> Idents(v$scData1)
VlnPlot(v$scData1, input$deg.gene)
}
})
}
})
output$Deg.plot <- renderPlotly({
plotViolin()
})
output$download_violn <- downloadHandler(
filename = function(){"Violin plot.png"},
content = function(fname){
ggsave(fname,plotViolin(), height = 7, width = 7)
}
)
plotFeature <- reactive({
if(is.null(v$scData1) || is.null(v$isVisdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg2 == "seurat_clusters"){
v$scData1$seurat_clusters -> Idents(v$scData1)
FeaturePlot(v$scData1, input$deg.gene)
}
else if(input$deg2 == "primary.predict"){
v$scData1$primary.predict -> Idents(v$scData1)
FeaturePlot(v$scData1, input$deg.gene)
}
else if(input$deg2 == "newID"){
v$scData1$newID -> Idents(v$scData1)
FeaturePlot(v$scData1, input$deg.gene)
}
})
}
})
output$Deg1.plot <- renderPlotly({
plotFeature()
})
output$download_feature <- downloadHandler(
filename = function(){"Feature plot.png"},
content = function(fname){
ggsave(fname,plotFeature(), height = 7, width = 7)
}
)
plotRidge <- reactive({
if(is.null(v$scData1) || is.null(v$isVisdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg2 == "seurat_clusters"){
v$scData1$seurat_clusters -> Idents(v$scData1)
RidgePlot(v$scData1, features = input$deg.gene)
}
else if(input$deg2 == "primary.predict"){
v$scData1$primary.predict -> Idents(v$scData1)
RidgePlot(v$scData1, features = input$deg.gene)
}
else if(input$deg2 == "newID"){
v$scData1$newID -> Idents(v$scData1)
RidgePlot(v$scData1, features = input$deg.gene)
}
})
}
})
output$Deg2.plot <- renderPlot({
plotRidge()
})
output$download_ridge <- downloadHandler(
filename = function(){"Ridge plot.png"},
content = function(fname){
ggsave(fname,plotRidge(), height = 7, width = 7)
}
)
output$Deg3.plot <- renderPlot({
if(is.null(v$ips.markers) || is.null(v$isDEGdone)){
plotly_empty()
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1, features = top10$gene, size = 3, angle = 30) + theme(axis.text.y = element_text(size = 4)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
output$Deg.table <- DT::renderDataTable(
v$ips.markers, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$gene1.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
if(input$deg3 == "seurat_clusters"){
selectInput("gene1", label = "Celltype1",
choices = as.vector(v$scData1$seurat_clusters))
}
else if(input$deg3 == "primary.predict"){
selectInput("gene1", label = "Celltype1",
choices = as.vector(v$scData1$primary.predict))
}
else if(input$deg3 == "newID"){
selectInput("gene1", label = "Celltype1",
choices = as.vector(v$scData1$newID))
}
}
})
output$gene2.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
if(input$deg3 == "seurat_clusters"){
selectInput("gene2", label = "Celltype2",
choices = as.vector(v$scData1$seurat_clusters))
}
else if(input$deg3 == "primary.predict"){
selectInput("gene2", label = "Celltype2",
choices = as.vector(v$scData1$primary.predict))
}
else if(input$deg3 == "newID"){
selectInput("gene2", label = "Celltype2",
choices = as.vector(v$scData1$newID))
}
}
})
observeEvent(input$doVolcano, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
v$scData1_subset <- subset(v$scData1, subset = primary.predict == input$gene1 | primary.predict == input$gene2)
if (input$norm1 == "LogNormalize"){
ips.markers_a <- FindAllMarkers(v$scData1_subset, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "RNA", test.use = input$test.use_a)
ips.markers_b <- FindMarkers(v$scData1_subset, ident.1 = input$gene1, ident.2 = input$gene2, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "RNA", test.use = input$test.use_a)
v$ips.markers_a <- ips.markers_a
v$ips.markers_b <- ips.markers_b
}
if (input$norm1 == "SCTransform"){
ips.markers_a <- FindAllMarkers(v$scData1_subset, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "SCT", test.use = input$test.use_a)
ips.markers_b <- FindMarkers(v$scData1_subset, ident.1 = input$gene1, ident.2 = input$gene2, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "SCT", test.use = input$test.use_a)
v$ips.markers_a <- ips.markers_a
v$ips.markers_b <- ips.markers_b
}
shinyalert("Pairwise DEGs done", "Pairwise DEGs done, please run GSEA Analysis", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$volcano.plot <- renderPlot({
if(is.null(v$ips.markers_b)){
plotly_empty()
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
EnhancedVolcano(toptable = v$ips.markers_b, lab = row.names(v$ips.markers_b), x ="avg_log2FC", y ="p_val_adj", pointSize = 1, labSize = 5, legendLabSize = 12, axisLabSize = 12)
})
}
})
output$dega.plot <- renderPlot({
if(is.null(v$ips.markers_a)){
plotly_empty()
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
v$ips.markers_a %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1_subset, features = top10$gene, size = 4, angle = 45) + theme(axis.text.y = element_text(size = 6)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
##---------------GSEA of scRNA-seq-------------------##
output$gsea.ct1.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
selectInput("gsea.ct1", label = "Celltype1",
choices = as.vector(v$scData1$primary.predict))
}
})
output$gsea.ct2.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
selectInput("gsea.ct2", label = "Celltype2",
choices = as.vector(v$scData1$primary.predict))
}
})
observeEvent(input$gsea, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating gene set enrichment analysis...", value=0, {
if(input$species_gsea == "Homo sapiens" & input$category_gsea == "H"){
v$msigdbr_hs_go <- msigdbr(species = "Homo sapiens", category = "H")
print(v$msigdbr_hs_go)
v$pathways <- split(x = v$msigdbr_hs_go$gene_symbol, f = v$msigdbr_hs_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Mus musculus" & input$category_gsea == "H"){
v$msigdbr_mm_go <- msigdbr(species = "Mus musculus", category = "H")
print(v$msigdbr_mm_go)
v$pathways <- split(x = v$msigdbr_mm_go$gene_symbol, f = v$msigdbr_mm_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Homo sapiens" & input$category_gsea == "C2"){
v$msigdbr_hs_go <- msigdbr(species = "Homo sapiens", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_hs_go)
v$pathways <- split(x = v$msigdbr_hs_go$gene_symbol, f = v$msigdbr_hs_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Mus musculus" & input$category_gsea == "C2"){
v$msigdbr_mm_go <- msigdbr(species = "Mus musculus", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_mm_go)
v$pathways <- split(x = v$msigdbr_mm_go$gene_symbol, f = v$msigdbr_mm_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Homo sapiens" & input$category_gsea == "C5"){
v$msigdbr_hs_go <- msigdbr(species = "Homo sapiens", category = "C5")
print(v$msigdbr_hs_go)
v$pathways <- split(x = v$msigdbr_hs_go$gene_symbol, f = v$msigdbr_hs_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Mus musculus" & input$category_gsea == "C5"){
v$msigdbr_mm_go <- msigdbr(species = "Mus musculus", category = "C5")
print(v$msigdbr_mm_go)
v$pathways <- split(x = v$msigdbr_mm_go$gene_symbol, f = v$msigdbr_mm_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Homo sapiens" & input$category_gsea == "C7"){
v$msigdbr_hs_go <- msigdbr(species = "Homo sapiens", category = "C7")
print(v$msigdbr_hs_go)
v$pathways <- split(x = v$msigdbr_hs_go$gene_symbol, f = v$msigdbr_hs_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Mus musculus" & input$category_gsea == "C7"){
v$msigdbr_mm_go <- msigdbr(species = "Mus musculus", category = "C7")
print(v$msigdbr_mm_go)
v$pathways <- split(x = v$msigdbr_mm_go$gene_symbol, f = v$msigdbr_mm_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Homo sapiens" & input$category_gsea == "C8"){
v$msigdbr_hs_go <- msigdbr(species = "Homo sapiens", category = "C8")
print(v$msigdbr_hs_go)
v$pathways <- split(x = v$msigdbr_hs_go$gene_symbol, f = v$msigdbr_hs_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
if(input$species_gsea == "Mus musculus" & input$category_gsea == "C8"){
v$msigdbr_mm_go <- msigdbr(species = "Mus musculus", category = "C8")
print(v$msigdbr_mm_go)
v$pathways <- split(x = v$msigdbr_mm_go$gene_symbol, f = v$msigdbr_mm_go$gs_name)
print(v$pathways)
v$markers <- FindMarkers(v$scData1, ident.1 = input$gsea.ct1, ident.2 = input$gsea.ct2, min.pct = input$min_pct1, logfc.threshold = input$logfc1, test.use = input$test.use1)
v$markers <- v$markers %>% arrange(desc(avg_log2FC))
print(v$markers)
v$markers.log2FC <- v$markers$avg_log2FC
names(v$markers.log2FC) <- row.names(v$markers)
v$markers.log2FC <- sort(na.omit(v$markers.log2FC), decreasing = TRUE)
print(v$markers.log2FC)
v$fgseaRes <- fgsea(pathways = v$pathways, stats = v$markers.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes)
v$topPathwaysUp <- v$fgseaRes[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown <- v$fgseaRes[ES < 0][head(order(pval), n=10), pathway]
v$topPathways <- c(v$topPathwaysUp, rev(v$topPathwaysDown))
}
output$gsea.done <- renderText(paste0("Gene set enrichment done!"))
v$isGSEAdone <- TRUE
})
}
})
output$gsea.select <- renderUI({
if(is.null(v$pathways)){
return(NULL)
}else{
selectInput("gsea.pathway", label = "Gene set to visualise",
choices = names(v$pathways))
}
})
output$gsea_plot <- renderPlot({
if(is.null(v$pathways) || is.null(v$isGSEAdone)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotEnrichment(v$pathways[[input$gsea.pathway]], v$markers.log2FC) + labs(title=input$gsea.pathway)
})
}
})
output$gsea_plot1 <- renderPlot({
if(is.null(v$pathways) || is.null(v$isGSEAdone)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotGseaTable(v$pathways[v$topPathways], v$markers.log2FC, v$fgseaRes, gseaParam=0.5)
})
}
})
output$gsea.table <- DT::renderDataTable(
v$fgseaRes, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_gsea.table <- downloadHandler(
filename = function(){"GSEA Results.csv"},
content = function(fname){
withProgress(message="Downloading GSEA Results...", value=0, {
fwrite(v$fgseaRes, fname)
})
}
)
##---------------Trajectory of scRNA-seq (not working)-------------------
observeEvent(input$doMonocle3, {
withProgress(message = "Running Monocle3...", value = 0.3, {
v$scDatad <- DietSeurat(v$scData, graphs = "umap")
v$scDatatr <- as.cell_data_set(v$scDatad)
v$scDatatr <- cluster_cells(cds = v$scDatatr, reduction_method = "UMAP")
v$scDatatr <- learn_graph(v$scDatatr, use_partition = TRUE)
cell <- WhichCells(v$scData, idents = 0)
v$scDatatr <- order_cells(v$scDatatr, reduction_method = "UMAP", root_cells = cell)
output$Trajectory.done <- renderText(paste0("Trajectory done!"))
v$isTrajectorydone <- TRUE
})
})
output$Monocle3_plot <- renderPlotly({
if(is.null(v$scData) || is.null(v$isTrajectorydone)){
plotly_empty()
}else{
withProgress(message="Generating Trajectory Plot...", value=0, {
plot_cells(cds = v$scDatatr, color_cells_by = "pseudotime", show_trajectory_graph = TRUE)
})
}
})
##---------------Cell-cell communication of scRNA-seq-------------------
observeEvent(input$doCC, {
tpmFiles <- v$scData1
if (is.null(tpmFiles)){
v$scData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running cell-cell communication...", value = 0.3, {
if(input$cc1a == "primary.predict"){
v$scData1$primary.predict -> Idents(v$scData1)
v$cellphone <- liana_wrap(v$scData1, method = input$cc_method, resource = input$cc_resource)
}
if(input$cc1a == "seurat_clusters"){
v$scData1$seurat_clusters -> Idents(v$scData1)
v$cellphone <- liana_wrap(v$scData1, method = input$cc_method, resource = input$cc_resource)
}
output$cc.done <- renderText(paste0("Cell-cell communication done!"))
v$isCCdone <- TRUE
shinyalert("Cell-cell communication done", "Cell-cell communication done", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$CC_plot1 <- renderPlot({
if(is.null(v$isCCdone)){
plotly_empty()
}else{
withProgress(message="Generating cell-cell communication plot...", value=0, {
heat_freq(v$cellphone, pallette = c("blue", "white", "red"))
#heatmaps_plot(meta_file = 'metadata.txt', pvalues_file = 'out/pvalues.txt', count_filename = 'heatmap_count.pdf', log_filename = 'heatmap_log_count.pdf', count_network_filename = 'count_network.txt', interaction_count_filename = 'interactions_count.txt', count_network_separator = '\t', interaction_count_separator = '\t', show_rownames = T, show_colnames = T, scale="none", cluster_cols = T,border_color='white', cluster_rows = T, fontsize_row=11, fontsize_col = 11, main = '',treeheight_row=0, family='Arial', treeheight_col = 0, meta_sep='\t', pvalues_sep='\t', pvalue=input$cc_pval)
})
}
})
output$CC.gene.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
selectInput("CC.gene1", label = "Sender",multiple = T,
choices = unique(v$scData1$primary.predict), selected = unique(v$scData1$primary.predict)[1])
}
})
output$CC.gene1.select <- renderUI({
if(is.null(v$scData1)){
plotly_empty()
}else{
selectInput("CC.gene2", label = "Reciever",multiple = T,
choices = unique(v$scData1$primary.predict), selected = unique(v$scData1$primary.predict)[2])
}
})
output$CC_plot2 <- renderPlot({
if(is.null(v$isCCdone)){
plotly_empty()
}else{
withProgress(message="Generating cell-cell communication plot...", value=0, {
chord_freq(v$cellphone, source_groups = input$CC.gene1, target_groups = input$CC.gene2)
})
}
})
output$cc.table <- DT::renderDataTable(
v$cellphone, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
#output$cc.table1 <- DT::renderDataTable(
# v$pvals, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
#observeEvent(input$loadexample_atacseq, {
# withProgress(message="Loading example Data...", value=0.5, {
# tpmFiles_example_atac <- Read10X_h5(filename = "atac_pbmc_5k_v1/filtered_peak_bc_matrix.h5")
# meta_example_atac <- read.csv('atac_pbmc_5k_v1/singlecell.csv', header = T, row.names = 1)
# chrom_assay_example_atac <- CreateChromatinAssay(counts = tpmFiles_example_atac, sep = c(":", "-"), genome = 'hg19', fragments = 'atac_pbmc_5k_v1/fragments.tsv.gz', min.cells = 10, min.features = 200)
# sObj_example_atac <- CreateSeuratObject(chrom_assay_example_atac, assay = "peaks", project = input$projName4, meta.data = meta_example_atac)
# print(sObj_example_atac)
# annotations_atac <- GetGRangesFromEnsDb(ensdb = EnsDb.Hsapiens.v86)
# seqlevelsStyle(annotations_atac) <- 'UCSC'
# genome(annotations_atac) <- "hg19"
# Annotation(sObj_example_atac) <- annotations_atac
# sObj_example_atac <- NucleosomeSignal(object = sObj_example_atac)
# sObj_example_atac <- TSSEnrichment(object = sObj_example_atac, fast = FALSE)
# sObj_example_atac$pct_reads_in_peaks <- sObj_example_atac$peak_region_fragments / sObj_example_atac$passed_filters * 100
# sObj_example_atac$blacklist_ratio <- sObj_example_atac$blacklist_region_fragments / sObj_example_atac$peak_region_fragments
# sObj_example_atac$high.tss <- ifelse(sObj_example_atac$TSS.enrichment > 2, 'High', 'Low')
# sObj_example_atac$nucleosome_group <- ifelse(sObj_example_atac$nucleosome_signal > 4, 'NS > 4', 'NS < 4')
# sObj_example_atac <- RunTFIDF(sObj_example_atac)
# sObj_example_atac <- FindTopFeatures(sObj_example_atac, min.cutoff = 'q0')
# sObj_example_atac <- RunSVD(sObj_example_atac)
# sObj_example_atac <- FindNeighbors(object = sObj_example_atac, reduction = 'lsi', dims = 2:30)
# sObj_example_atac <- FindClusters(object = sObj_example_atac, resolution = 0.5, verbose = FALSE, algorithm = 3)
# sObj_example_atac <- RunUMAP(object = sObj_example_atac, reduction = 'lsi', dims = 2:30)
# gene.activities <- GeneActivity(sObj_example_atac, features = VariableFeatures(v$scData1))
# sObj_example_atac[["ACTIVITY"]] <- CreateAssayObject(counts = gene.activities)
# DefaultAssay(sObj_example_atac) <- "ACTIVITY"
# sObj_example_atac <- NormalizeData(sObj_example_atac)
# sObj_example_atac <- ScaleData(sObj_example_atac, features = rownames(sObj_example_atac))
# transfer.anchors <- FindTransferAnchors(reference = v$scData1, query = sObj_example_atac, features = VariableFeatures(object = v$scData1), reference.assay = "RNA", query.assay = "ACTIVITY", reduction = "cca")
# celltype.predictions <- TransferData(anchorset = transfer.anchors, refdata = v$scData1$primary.predict, weight.reduction = sObj_example_atac[["lsi"]], dims = 2:30)
# sObj_example_atac <- AddMetaData(sObj_example_atac, metadata = celltype.predictions)
# v$transfer.anchors <- transfer.anchors
# v$atacData <- sObj_example_atac
# })
# label1 <- "Example loaded"
# updateActionButton(inputId = "loadexample_atacseq", label = label1)
# shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
#})
observeEvent(input$annotate_atacseq, {
#tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Annotating...", value=0.5, {
transfer.anchors <- FindTransferAnchors(reference = v$scData1, query = v$atacData, reduction = "cca")
celltype.predictions <- TransferData(anchorset = transfer.anchors, refdata = v$scData1$primary.predict, weight.reduction = v$atacData[["lsi"]], dims = 2:30)
print(celltype.predictions)
v$atacData <- AddMetaData(v$atacData, metadata = celltype.predictions)
print("res1")
print(v$atacData@meta.data)
v$transfer.anchors <- transfer.anchors
})
label1 <- "Annotation done"
updateActionButton(inputId = "loadexample_atacseq", label = label1)
shinyalert("Annotation done", "Annotation done.", type = "success", imageWidth = 10, imageHeight = 10)
v$isCELLAtacsdone <- TRUE
}
})
observeEvent(input$annotate_atacseq_a, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Annotating...", value=0.5, {
transfer.anchors1 <- FindTransferAnchors(reference = v$scData1, query = v$atacData, reduction = "cca")
celltype.predictions1 <- TransferData(anchorset = transfer.anchors1, refdata = v$scData1$primary.predict, weight.reduction = v$atacData[["lsi"]], dims = 2:30)
v$atacData <- AddMetaData(v$atacData, metadata = celltype.predictions1)
v$transfer.anchors1 <- transfer.anchors1
})
label1 <- "Annotation done"
updateActionButton(inputId = "loadexample_atacseq1", label = label1)
shinyalert("Annotation done", "Annotation done.", type = "success", imageWidth = 10, imageHeight = 10)
}
})
observeEvent(input$annotate1_atacseq_a, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Annotating...", value=0.5, {
transfer.anchors1 <- FindTransferAnchors(reference = v$scData1, query = v$atacData, reduction = "cca")
celltype.predictions1 <- TransferData(anchorset = transfer.anchors1, refdata = v$scData1$primary.predict, weight.reduction = v$atacData[["lsi"]], dims = 2:30)
v$atacData <- AddMetaData(v$atacData, metadata = celltype.predictions1)
v$transfer.anchors1 <- transfer.anchors1
})
label1 <- "Annotation done"
updateActionButton(inputId = "loadexample_atacseq1", label = label1)
shinyalert("Annotation done", "Annotation done.", type = "success", imageWidth = 10, imageHeight = 10)
}
})
#output$annotate_scRNA_ATAC_plot <- renderPlot({
# if(is.null(v$scData1)|| is.null(v$atacData)){
# return(NULL)
# }else{
# withProgress(message="Generating ATAC UMAP...", value=0, {
# DimPlot(v$atacData, group.by = "seurat_clusters", label = T)
# })
# }
#})
#output$annotate_scRNA_ATAC_plot1 <- renderPlot({
# if(is.null(v$scData1)|| is.null(v$atacData)){
# return(NULL)
# }else{
# withProgress(message="Generating ATAC UMAP...", value=0, {
# DimPlot(v$atacData, group.by = "predicted.id", label = T)
# })
# }
#})
#observe({
# if(input$process_atacseq > 0){
# print('2')
# session$sendCustomMessage("myCallbackHandler1", "2")
# }
#})
observeEvent(input$annotate1_atacseq, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Annotating...", value=0.5, {
transfer.anchors <- FindTransferAnchors(reference = v$scData1, query = v$atacData, reduction = "cca")
celltype.predictions <- TransferData(anchorset = transfer.anchors, refdata = v$scData1$primary.predict, weight.reduction = v$atacData[["lsi"]], dims = 2:30)
v$atacData <- AddMetaData(v$atacData, metadata = celltype.predictions)
v$transfer.anchors <- transfer.anchors
})
label1 <- "Annotation done"
updateActionButton(inputId = "loadexample_atacseq", label = label1)
shinyalert("Annotation done", "Annotation done.", type = "success", imageWidth = 10, imageHeight = 10)
}
})
##------------------------Data integration module--------------------------##
output$integration_image <- renderImage({
list(src = "www/Picture3.png",
height = 500)
}, deleteFile = FALSE)
observeEvent(input$loadexample1, {
withProgress(message="Loading example Data...", value=0.5, {
tpmFiles1 <- read.table("integration/concatenated_expr_data.txt", header = T, row.names = 1, check.names = F, sep = '\t')
#scH5 <- input$scH5
annoFile1 <- read.table("integration/metadata.txt", header = T, row.names = 1, sep = '\t')
})
if (is.null(tpmFiles1)){
v$scData2 <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFiles1)
#print(tpmFiles$name)
print(file.exists(paste(tpmFiles1$datapath[1], "/", tpmFiles1$name[1], sep="")))
v$tpmFiles1 <- tpmFiles1
v$anno <- annoFile1
sObj1 <- CreateSeuratObject(v$tpmFiles1,
meta.data = v$anno,
project = input$projName1,
min.genes = input$min.genes1,
min.cells = input$min.cells1)
sObj1[["percent.mt"]] <- PercentageFeatureSet(sObj1, pattern = "^MT-")
v$scData2 <- sObj1
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample1", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$loadexample1a, {
withProgress(message="Loading example Data...", value=0.5, {
data1 <- Read10X_h5("integration/filtered_feature_bc_matrix_data1.h5")
data2 <- Read10X_h5("integration/filtered_feature_bc_matrix_data2.h5")
})
if (is.null(data1) || is.null(data2)){
v$scData1 <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(data1)
print(data2)
#print(tpmFiles$name)
data <- list(data1, data2)
for (i in 1:length(data)){
data[[i]] <- CreateSeuratObject(as.matrix(data[[i]]), project = input$projName1, min.genes = input$min.genes1, min.cells = input$min.cells1)
}
sObj1 <- merge(data[[1]], data[[2]])
sObj1[["percent.mt"]] <- PercentageFeatureSet(sObj1, pattern = "^MT-")
sObj1[["batch"]] <- ifelse(endsWith(sObj1@assays$RNA@data@Dimnames[[2]], "1"), "1", "2")
v$scData2 <- sObj1
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample1a", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$loadButton1, {
if(input$scInput1 == "Raw Counts Matrix"){
tpmFiles1 <- input$tpmFiles1
annoFile1 <- input$cellAnnoFiles1
names.field <- input$field
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFiles1$datapath)
print(tpmFiles1$name)
print(file.exists(paste(tpmFiles1$datapath[1], "/", tpmFiles1$name[1], sep="")))
file_list <- tpmFiles1$datapath
names(file_list) <- basename(file_list)
exp.data1 <- imap(file_list, function(x, y) {
cts <- x %>%
read.table (header = T, row.names = 1, stringsAsFactors = F, check.names = F) %>%
CreateSeuratObject(project=y)
})
print(exp.data1)
exp.data2 <- merge(exp.data1[[1]], exp.data1[2:length(exp.data1)])
if(!is.null(annoFile1)){
anno.data1 <- rbindlist(lapply(annoFile1$datapath, fread), use.names = TRUE, fill = TRUE)
anno.data2 <- data.frame(anno.data1, row.names = 1)
}
print(anno.data2)
incProgress(0.5, "Creating Seurat Object")
sObj1 <- AddMetaData(exp.data2, anno.data2)
sObj1$orig.ident <- "Integration"
Idents(sObj1) <- sObj1$orig.ident
sObj1[["percent.mt"]] <- PercentageFeatureSet(sObj1, pattern = "^MT-")
print(sObj1@meta.data)
v$scData2 <- sObj1
print(v$scData2@meta.data)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput1 == "10X cellranger"){
scH5_1 <- input$scH5_1
if (is.null(scH5_1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(scH5_1$datapath)
print(scH5_1$name)
print(file.exists(paste(scH5_1$datapath[1], "/", scH5_1$name[1], sep="")))
file_list <- scH5_1$datapath
names(file_list) <- basename(file_list)
exp.data1 <- imap(file_list, function(x, y) {
cts <- x %>%
Read10X_h5 %>%
CreateSeuratObject(project=y)
})
print(exp.data1)
exp.data2 <- merge(exp.data1[[1]], exp.data1[2:length(exp.data1)])
print(exp.data2)
incProgress(0.5, "Creating Seurat Object")
exp.data2$orig.ident <- "Integration"
Idents(exp.data2) <- exp.data2$orig.ident
exp.data2[["percent.mt"]] <- PercentageFeatureSet(exp.data2, pattern = "^MT-")
exp.data2[["batch"]] <- ifelse(endsWith(exp.data2@assays$RNA@data@Dimnames[[2]], "1"), "1", "2")
v$scData2 <- exp.data2
print(v$scData2@meta.data)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
})
output$countdataDT1 <- renderDataTable({
if(!is.null(v$scData2))
{
if(ncol(v$scData2) > 20 )
return(as.matrix(v$scData2@assays$RNA@counts[,1:20]))
}
}, server = FALSE)
observeEvent(input$filter_seurat1, {
if(input$scInput1 == "Raw Counts Matrix"){
tpmFiles1 <- input$tpmFiles1
annoFile1 <- input$cellAnnoFiles1
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scData2)
v$scData2 <- subset(v$scData2, subset = nFeature_RNA > input$ob1a & nFeature_RNA < input$ob2a & percent.mt < input$ob3a)
#sObj1[["batch"]] <- ifelse(endsWith(sObj1@assays$RNA@data@Dimnames[[2]], "1"), "1", "2")
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
else if(input$scInput1 == "10X cellranger"){
scH5_1 <- input$scH5_1
if (is.null(scH5_1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scData2)
sObj1 <- subset(v$scData2, subset = nFeature_RNA > input$ob1a & nFeature_RNA < input$ob2a & percent.mt < input$ob3a)
sObj1$orig.ident <- "Integration"
Idents(sObj1) <- sObj1$orig.ident
#sObj1[["percent.mt"]] <- PercentageFeatureSet(sObj1, pattern = "^MT-")
v$scData2 <- sObj1
print(v$scData2@meta.data)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
})
observeEvent(input$reset_intg, {
session$reload()
print("Reset done")
})
output$nFeature_RNAPlot1 <- renderPlot({
if(is.null(v$scData2)){
plotly_empty()
}else{
VlnPlot(v$scData2, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3) + NoLegend()
}
})
output$mitoPlot1 <- renderPlot({
if(is.null(v$scData2)){
plotly_empty()
}else{
VlnPlot(v$scData2, "percent.mt") + NoLegend()
}
})
output$nCount_RNAPlot1 <- renderPlot({
if(is.null(v$scData2)){
plotly_empty()
}else{
VlnPlot(v$scData2, "nCount_RNA") + NoLegend()
}
})
output$FeatureScatterPlot1a <- renderPlotly({
if(is.null(v$scData2)){
plotly_empty()
}else{
print(FeatureScatter(v$scData2, "nCount_RNA", "nFeature_RNA"))
}
})
output$FeatureScatterPlot2a <- renderPlotly({
if(is.null(v$scData2)){
plotly_empty()
}else{
print(FeatureScatter(v$scData2, "nCount_RNA", "percent.mt"))
}
})
#observe({if(input$scAnalysis_integ == "Seurat" || input$scAnalysis_integ == "Harmony" || input$scAnalysis_integ == "fastMNN" || input$scAnalysis_integ == "scVI"){
observeEvent(input$findVarGenes_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding variable genes...", value = 0, {
v$scData2 <- NormalizeData(v$scData2)
v$scData2 <- FindVariableFeatures(v$scData2,
mean.function = ExpMean,
dispersion.function = LogVMR,
nfeatures = input$var.genes_bef_intg,
selection.method = input$selection.method)
#all.genes <- rownames(v$scData1)
v$scData2 <- ScaleData(v$scData2)
print(v$scData2)
incProgress(0.5)
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData1@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scData2)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scData2), 10)
variable_feature1 <- VariableFeaturePlot(v$scData2)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("Highly variable features identified", "Highly variable features identified, please perform PCA (Before Integration) ", type = "success", imageWidth = 10, imageHeight = 10)
#dev.off()
})
}
}
output$VarGenes_bef_intg <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scData2)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Seurat_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scData2)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scData2@assays$RNA@var.features, file = txtfile)
})
}
})
})
}
})
observeEvent(input$runPCA_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0,{
incProgress(0.5, message = "Running PCA...")
if (input$scInput1 == "Raw Counts Matrix")
{
v$scData2 <- RunPCA(v$scData2, verbose = FALSE)
print(v$scData2[["pca"]], dims = 1:5, nfeatures = 5)
v$isPCAdone_bef_intg <- TRUE
PCA_plot1a <- DimPlot(v$scData2, reduction = "pca", label = T, label.size = 3)
PCA_plot1b <- DimPlot(v$scData2, reduction = "pca", label = T, group.by = 'batch', label.size = 3)
PCA_plot1c <- DimPlot(v$scData2, reduction = "pca", label = T, group.by = 'celltype', label.size = 3)
print(PCA_plot1a)
print(PCA_plot1b)
print(PCA_plot1c)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scData2)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
shinyalert("PCA performed", "PCA performed, please perform clustering", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger")
{
v$scData2 <- RunPCA(v$scData2, verbose = FALSE)
print(v$scData2@meta.data)
print(v$scData2[["pca"]], dims = 1:5, nfeatures = 5)
v$isPCAdone_bef_intg <- TRUE
PCA_plot1a <- DimPlot(v$scData2, reduction = "pca", label = T, label.size = 3)
PCA_plot1b <- DimPlot(v$scData2, reduction = "pca", label = T, group.by = 'batch', label.size = 3)
print(PCA_plot1a)
print(PCA_plot1b)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scData2)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
shinyalert("PCA performed", "PCA performed, please perform clustering", type = "success", imageWidth = 10, imageHeight = 10)
}
}
)
output$PCAplot_bef_tpm1 <- renderPlotly({
if(is.null(v$isPCAdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "pca", label = T, label.size = 3)
})
}
})
output$PCAplot_bef_tpm2 <- renderPlotly({
if(is.null(v$isPCAdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "pca", label = T, group.by = 'batch', label.size = 3)
})
}
})
output$PCAplot_bef_tpm3 <- renderPlotly({
if(is.null(v$isPCAdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "pca", label = T, group.by = 'celltype', label.size = 3)
})
}
})
output$PCAplot_bef_h5_1 <- renderPlotly({
if(is.null(v$isPCAdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "pca", label = T, label.size = 3)
})
}
})
output$PCAplot_bef_h5_2 <- renderPlotly({
if(is.null(v$isPCAdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "pca", label = T, group.by = 'batch', label.size = 3)
})
}
})
output$vizPlot_bef_intg <- renderPlot({
if(is.null(v$scData2)){
plotly_empty()
}else{
VizDimLoadings(v$scData2, dims = as.numeric(input$select.pc_bef_intg))
}
})
output$PCHeatmap_bef_intg <- renderPlot({
if(is.null(v$scData2)){
plotly_empty()
}else{
DimHeatmap(v$scData2, dims = as.numeric(input$select.pc_bef_intg))
}
})
output$PCtable_bef_intg <- DT::renderDataTable({
if(is.null(v$scData2) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_bef_intg <- renderPlot({
if(is.null(v$scData2)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scData2, ndims = 50)
})
}
}, height = 400, width = 450)
observeEvent(input$findCluster_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
v$scData2 <- FindNeighbors(v$scData2, reduction = "pca", dims = 1:input$dim.used_bef_intg)
v$scData2 <- FindClusters(v$scData2, resolution = input$clus.res_bef_intg)
#output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone1 <- TRUE
shinyalert("Clustering performed", "Clustering performed, please perform UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Cluster2DPlot_bef_intg <- renderPlotly({
if(is.null(v$isClusterdone1)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scData2, reduction = "umap", label = T, label.size = 3)
})
}
})
observeEvent(input$findoptimumCluster1, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding optimum resolution...", value = 0.3, {
#if(input$assays1 == "LogNormalization"){
#DefaultAssay(v$scData1) <- "RNA"
reses<-seq(from = input$clus.res_a1, to = input$clus.res_b1, by = 0.1)
for (res in reses){
v$scData2<-FindClusters(v$scData2, resolution = res)
nores<-gsub(pattern = ".", replacement = "", res, fixed = T)
}
print(v$scData2@meta.data)
output$optimumcluster.done <- renderText(paste0("Clustering done!"))
#}
#else if(input$assays1 == "SCTransform"){
# DefaultAssay(v$scData1) <- "SCT"
# v$scData1 <- FindNeighbors(v$scData1, dims = 1:input$dim.used, assay = "SCT", nn.method = "rann")
# v$scData1 <- FindClusters(v$scData1, resolution = input$clus.res)
# output$cluster.done <- renderText(paste0("Clustering done!"))
#}
v$isOptimumClusterdone1 <- TRUE
shinyalert("Number of clusters identified at different resolution", "Number of clusters identified at different resolution, please perform UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
plotOptimumCluster1 <- reactive({
if(is.null(v$scData2) || is.null(v$isOptimumClusterdone1)){
plotly_empty()
}else{
withProgress(message="Plotting Clustering tree...", value=0, {
p <- clustree(v$scData2)
})
}
})
output$OptimumCluster2DPlot_a1 <- renderPlot({
print(plotOptimumCluster1())
})
output$download_OptimumCluster1 <- downloadHandler(
filename = function(){"Clustree plot.png"},
content = function(fname){
ggsave(fname,plotOptimumCluster1(), height = 7, width = 7)
}
)
output$download_OptimumClusterTable1 <- downloadHandler(
filename = function(){"Clustering_different_resolution.csv"},
content = function(fname){
withProgress(message="Downloading Cluster Table at different resolution...", value=0, {
write.csv(v$scData2@meta.data, fname)
})
}
)
output$subcluster.gene.select1 <- renderUI({
if(is.null(v$scData2)){
plotly_empty()
}else{
selectInput("cluster_subcluster1", label = "Cluster to subcluster",
choices = unique(v$scData2$seurat_clusters), selected = unique(v$scData2$seurat_clusters)[1])
}
})
observeEvent(input$subcluster1, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Performing subcluster analysis...", value=0, {
v$scData2$seurat_clusters -> Idents(v$scData2)
v$scData2.subset <- FindSubCluster(v$scData2, input$cluster_subcluster1, "RNA_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res1, algorithm = 1)
print(v$scData2.subset)
output$Subcluster1.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone1 <- TRUE
})
}
})
plotSubclster1 <- reactive({
if(is.null(v$scData2.subset) || is.null(v$isSubclusterdone1)){
plotly_empty()
}else{
withProgress(message="Generating Celltype similarity plot...", value=0, {
p <- DimPlot(v$scData2.subset, reduction = "umap", group.by = "sub.cluster", label = T, label.size = 2.5)
})
}
})
output$subcluster_plot1 <- renderPlotly({
plotSubclster1()
})
output$download_subcluster1 <- downloadHandler(
filename = function(){"Subcluster plot.png"},
content = function(fname){
ggsave(fname,plotSubclster1(), height = 7, width = 7)
}
)
observeEvent(input$runUMAP_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0,{
incProgress(0.5, message = "Running UMAP...")
if (input$scInput1 == "Raw Counts Matrix")
{
v$scData2 <- RunUMAP(v$scData2, reduction = "pca", dims = 1:input$dim.used_bef_intg)
v$isUMAPdone_bef_intg <- TRUE
UMAP_plot1a <- DimPlot(v$scData2, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData2, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
UMAP_plot1c <- DimPlot(v$scData2, reduction = "umap", label = T, group.by = 'celltype', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
shinyalert("UMAP performed", "UMAP performed, please perform tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger")
{
v$scData2 <- RunUMAP(v$scData2, reduction = "pca", dims = 1:30)
v$isUMAPdone_bef_intg <- TRUE
UMAP_plot1a <- DimPlot(v$scData2, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData2, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
shinyalert("UMAP performed", "UMAP performed, please perform tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
output$UMAPplot_bef_tpm1 <- renderPlotly({
if(is.null(v$isUMAPdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "umap", label = T, label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_bef_tpm2 <- renderPlotly({
if(is.null(v$isUMAPdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "umap", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_bef_tpm3 <- renderPlotly({
if(is.null(v$isUMAPdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "umap", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_bef_h5_1 <- renderPlotly({
if(is.null(v$isUMAPdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "umap", label = T, label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_bef_h5_2 <- renderPlotly({
if(is.null(v$isUMAPdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "umap", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
observeEvent(input$runTSNE_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running TSNE...", value = 0,{
incProgress(0.5, message = "Running TSNE...")
if (input$scInput1 == "Raw Counts Matrix")
{
v$scData2 <- RunTSNE(v$scData2, reduction = "pca", dims = 1:input$dim.used_bef_intg)
v$isTSNEdone_bef_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData2, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData2, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
TSNE_plot1c <- DimPlot(v$scData2, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
print(TSNE_plot1a)
print(TSNE_plot1b)
print(TSNE_plot1c)
shinyalert("tSNE performed", "tSNE performed, please perform celltype annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger")
{
v$scData2 <- RunTSNE(v$scData2, reduction = "pca", dims = 1:input$dim.used_bef_intg)
v$isTSNEdone_bef_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData2, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData2, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
shinyalert("tSNE performed", "tSNE performed, please perform celltype annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
output$TSNEplot_bef_tpm1 <- renderPlotly({
if(is.null(v$isTSNEdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "tsne", label = T, label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_bef_tpm2 <- renderPlotly({
if(is.null(v$isTSNEdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "tsne", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_bef_tpm3 <- renderPlotly({
if(is.null(v$isTSNEdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_bef_h5_1 <- renderPlotly({
if(is.null(v$isTSNEdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "tsne", label = T, label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_bef_h5_2 <- renderPlotly({
if(is.null(v$isTSNEdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData2, reduction = "tsne", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
observeEvent(input$doCELLiD_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running CELLiD...", value = 0.3, {
ref = readRDS(url("https://www.immunesinglecell.org/api/vishuo/download/getCellidRef"))
v$scData2.rna.data.average = AverageExpression(v$scData2)
v$scData2.rna.data.average = round(v$scData2.rna.data.average$RNA, 2)
#v$scData2.rna.data.average = data.frame(v$scData2.rna.data.average$RNA)
if(input$cellatlas1 == "all"){
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, ref)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, adipose1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, adrenal_gland1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, blood1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, bone_marrow1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, brain1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, breast1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, breast_milk1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, eye1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, gut1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, heart1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, kidney1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, liver1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, lung1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, pancreas1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, PDAC1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, skin1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, testis1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, thymus1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1 == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
v$res_bef_intg = FastIntegration::CELLiD(v$scData2.rna.data.average, tonsil1)
print(v$res_bef_intg)
v$scData2$primary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),1]
v$scData2$secondary.predict = v$res_bef_intg[as.numeric(v$scData2$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_bef_intg) <- newheaders
print(v$scData2@meta.data)
output$CELLiD1.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_bef_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
v$isCELLiDdone_bef_intg <- TRUE
shinyalert("Celltype annotation performed", "Celltype annotation performed, please perform data integration", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_cellid_bef_intg <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isCELLiDdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scData2, reduction = "umap", group.by = "primary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_cellid_bef_intg1 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isCELLiDdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scData2, reduction = "umap", group.by = "secondary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$ct_bef_intg.table <- DT::renderDataTable(
v$res_bef_intg, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
observeEvent(input$doCelltypist_bef_intg, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Celltypist...", value = 0.3, {
sc <- reticulate::import("scanpy", convert = FALSE)
ct <- reticulate::import("celltypist", convert = FALSE)
sceasy::convertFormat(v$scData2, from = "seurat", to = "anndata", outFile = 'ct_scrna.h5ad')
v$adata = sc$read_h5ad('ct_scrna.h5ad')
v$res = ct$annotate(filename = 'ct_scrna.h5ad', model = input$celltypistatlas1, majority_voting=T)
print(v$res)
v$adata = v$res$to_adata()
print("fff")
v$adata$obs$to_csv('celltypist_predict.csv')
v$meta1 <- read.csv('celltypist_predict.csv', header = T, row.names = 1)
v$scData2 <- AddMetaData(v$scData2, metadata = v$meta1)
v$scData2$primary.predict <- v$scData2$majority_voting
v$scData2$secondary.predict <- v$scData2$predicted_labels
print(v$scData2@meta.data)
v$isCelltypistdone_bef_intg <- TRUE
shinyalert("Celltype annotation performed", "Celltype annotation performed, please perform cell-cell similarity", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_celltypist_bef_intg <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isCelltypistdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scData2, reduction = "umap", group.by = "majority_voting", label = T, label.size = 3, repel = T) + NoLegend()
})
}
})
output$Umap_celltypist_bef_intg1 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isCelltypistdone_bef_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scData2, reduction = "umap", group.by = "predicted_labels", label = T, label.size = 3) + NoLegend()
})
}
})
output$ct_celltypist_bef_intg.table <- DT::renderDataTable(
v$meta1, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
observeEvent(input$doIntg_seurat, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Data Integration...", value = 0.3, {
v$scData1.list <- SplitObject(v$scData2, split.by = "batch")
print(v$scData2)
print(v$scData1.list)
features <- SelectIntegrationFeatures(object.list = v$scData1.list, nfeatures = input$nfeatures_intg_seurat)
print(features)
v$scData1.anchors <- FindIntegrationAnchors(object.list = v$scData1.list, anchor.features = features)
v$scData1.combined <- IntegrateData(anchorset = v$scData1.anchors)
print(v$scData1.anchors)
print(v$scData1.combined)
#v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg_seurat)
#v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg_seurat)
print(v$scData1.combined)
v$isIntgSeuratdone <- TRUE
shinyalert("Data integration done", "Data integration done, please run PCA", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$runPCA_intg_seurat, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- ScaleData(v$scData1.combined, verbose = FALSE)
v$scData1.combined <- RunPCA(v$scData1.combined, verbose = FALSE)
#output$pca1.done <- renderText(paste0("PCA done!"))
v$isPCAdone_intg_seurat <- TRUE
shinyalert("PCA done", "PCA done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$PCAplot_seurat_tpm1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_seurat)){
return(NULL)
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "pca", label = T, label.size = 3) + NoLegend()
})
}
})
output$PCAplot_seurat_tpm2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_seurat)){
return(NULL)
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "pca", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$PCAplot_seurat_tpm3 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_seurat)){
return(NULL)
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "pca", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$PCAplot_seurat_h5_1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_seurat)){
return(NULL)
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "pca", label = T, label.size = 3) + NoLegend()
})
}
})
output$PCAplot_seurat_h5_2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_seurat)){
return(NULL)
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "pca", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$vizPlot_intg_seurat <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
VizDimLoadings(v$scData1.combined, dims = as.numeric(input$select.pc_intg_seurat))
}
})
output$PCHeatmap_intg_seurat <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
DimHeatmap(v$scData1.combined, dims = as.numeric(input$select.pc_intg_seurat))
}
})
output$PCtable_intg_seurat <- DT::renderDataTable({
if(is.null(v$scData1.combined) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_intg_seurat <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scData1.combined, ndims = 50)
})
}
}, height = 400, width = 450)
observeEvent(input$findCluster_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Seurat")
{
DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Harmony")
{
DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "scVI")
{
#DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "scvi", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "fastMNN")
{
#DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Seurat")
{
DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Harmony")
{
DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "scVI")
{
#DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "scvi", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "fastMNN")
{
#DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg2)
v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone_intg <- TRUE
shinyalert("Clustering done", "Clustering done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
output$Cluster2DPlot_intg <- renderPlotly({
if(is.null(v$isClusterdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", label = T) + NoLegend()
})
}
})
observeEvent(input$findoptimumCluster2, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding optimum resolution...", value = 0.3, {
#if(input$assays1 == "LogNormalization"){
#DefaultAssay(v$scData1) <- "RNA"
reses<-seq(from = input$clus.res_a2, to = input$clus.res_b2, by = 0.1)
for (res in reses){
v$scData1.combined<-FindClusters(v$scData1.combined, resolution = res)
nores<-gsub(pattern = ".", replacement = "", res, fixed = T)
}
print(v$scData1.combined@meta.data)
output$optimumcluster2.done <- renderText(paste0("Clustering done!"))
#}
#else if(input$assays1 == "SCTransform"){
# DefaultAssay(v$scData1) <- "SCT"
# v$scData1 <- FindNeighbors(v$scData1, dims = 1:input$dim.used, assay = "SCT", nn.method = "rann")
# v$scData1 <- FindClusters(v$scData1, resolution = input$clus.res)
# output$cluster.done <- renderText(paste0("Clustering done!"))
#}
v$isOptimumClusterdone2 <- TRUE
shinyalert("Number of clusters identified at different resolution", "Number of clusters identified at different resolution, please perform UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
plotOptimumCluster2 <- reactive({
if(is.null(v$scData1.combined) || is.null(v$isOptimumClusterdone2)){
plotly_empty()
}else{
withProgress(message="Plotting Clustering tree...", value=0, {
p <- clustree(v$scData1.combined)
})
}
})
output$OptimumCluster2DPlot_b1 <- renderPlot({
print(plotOptimumCluster2())
})
output$download_OptimumCluster2 <- downloadHandler(
filename = function(){"Clustree plot.png"},
content = function(fname){
ggsave(fname,plotOptimumCluster2(), height = 7, width = 7)
}
)
output$download_OptimumClusterTable2 <- downloadHandler(
filename = function(){"Clustering_different_resolution.csv"},
content = function(fname){
withProgress(message="Downloading Cluster Table at different resolution...", value=0, {
write.csv(v$scData1.combined@meta.data, fname)
})
}
)
output$subcluster.gene.select2 <- renderUI({
if(is.null(v$scData1.combined)){
plotly_empty()
}else{
selectInput("cluster_subcluster2", label = "Cluster to subcluster",
choices = unique(v$scData1.combined$seurat_clusters), selected = unique(v$scData1.combined$seurat_clusters)[1])
}
})
observeEvent(input$subcluster2, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Performing subcluster analysis...", value=0, {
if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Seurat")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "integrated_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Harmony")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "integrated_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "scVI")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "RNA_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "fastMNN")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "RNA_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Seurat")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "integrated_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Harmony")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "integrated_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "scVI")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "RNA_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "fastMNN")
{
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.subset <- FindSubCluster(v$scData1.combined, input$cluster_subcluster2, "RNA_snn", subcluster.name = "sub.cluster", resolution = input$subcluster.res2, algorithm = 1)
print(v$scData1.combined.subset)
output$Subcluster2.done <- renderText(paste0("Subclustering done!"))
v$isSubclusterdone2 <- TRUE
}
})
}
})
plotSubclster2 <- reactive({
if(is.null(v$scData1.combined.subset) || is.null(v$isSubclusterdone2)){
plotly_empty()
}else{
withProgress(message="Generating Celltype similarity plot...", value=0, {
p <- DimPlot(v$scData1.combined.subset, reduction = "umap", group.by = "sub.cluster", label = T, label.size = 2.5)
})
}
})
output$subcluster_plot2 <- renderPlotly({
plotSubclster2()
})
output$download_subcluster2 <- downloadHandler(
filename = function(){"Subcluster plot.png"},
content = function(fname){
ggsave(fname,plotSubclster2(), height = 7, width = 7)
}
)
observeEvent(input$runUMAP_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0,{
incProgress(0.5, message = "Running UMAP...")
if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Seurat")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
UMAP_plot1c <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'celltype', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Harmony")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
UMAP_plot1c <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'celltype', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "scVI")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "scvi", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
UMAP_plot1c <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'celltype', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "fastMNN")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
UMAP_plot1c <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'celltype', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Seurat")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Harmony")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "scVI")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "scvi", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "fastMNN")
{
v$scData1.combined <- RunUMAP(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg, spread = 1)
UMAP_plot1a <- DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3)
UMAP_plot1b <- DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3)
print(UMAP_plot1a)
print(UMAP_plot1b)
v$isUMAPdone_intg <- TRUE
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
output$UMAPplot_intg_tpm1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isUMAPdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_intg_tpm2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isUMAPdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_intg_tpm3 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isUMAPdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_intg_h5_1 <- renderPlot({
if(is.null(v$scData1.combined) || is.null(v$isUMAPdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", label = T, label.size = 3) + NoLegend()
})
}
})
output$UMAPplot_intg_h5_2 <- renderPlot({
if(is.null(v$scData1.combined) || is.null(v$isUMAPdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$UMAP_lisi_intg <- renderText({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Seurat")
{
print(v$scData1.combined@reductions$pca@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$pca@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Harmony")
{
print(v$scData1.combined@reductions$harmony@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$harmony@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "scVI")
{
print(v$scData1.combined@reductions$scvi@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$scvi@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "fastMNN")
{
print(v$scData1.combined@reductions$mnn@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$mnn@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Seurat")
{
print(v$scData1.combined@reductions$pca@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$pca@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Harmony")
{
print(v$scData1.combined@reductions$harmony@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$harmony@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "scVI")
{
print(v$scData1.combined@reductions$scvi@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$scvi@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "fastMNN")
{
print(v$scData1.combined@reductions$mnn@cell.embeddings)
print(v$scData1.combined@meta.data)
v$test <- paste("LISI Score:", median(lisi::compute_lisi(v$scData1.combined@reductions$mnn@cell.embeddings, v$scData1.combined@meta.data, 'batch')$batch))
print(v$test)
}
})
}
})
observeEvent(input$runTSNE_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running TSNE...", value = 0,{
incProgress(0.5, message = "Running TSNE...")
if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Seurat")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
TSNE_plot1c <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
print(TSNE_plot1c)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "Harmony")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
TSNE_plot1c <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
print(TSNE_plot1c)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "scVI")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "scvi", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
TSNE_plot1c <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
print(TSNE_plot1c)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "Raw Counts Matrix" & input$scAnalysis_integ == "fastMNN")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
TSNE_plot1c <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
print(TSNE_plot1c)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Seurat")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "pca", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "Harmony")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "scVI")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "scvi", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
else if (input$scInput1 == "10X cellranger" & input$scAnalysis_integ == "fastMNN")
{
v$scData1.combined <- RunTSNE(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg1)
v$isTSNEdone_intg <- TRUE
TSNE_plot1a <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3)
TSNE_plot1b <- DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3)
print(TSNE_plot1a)
print(TSNE_plot1b)
v$isTSNEdone_intg <- TRUE
shinyalert("tSNE done", "tSNE done, please perform cell type annotation", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
output$TSNEplot_intg_tpm1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isTSNEdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_intg_tpm2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isTSNEdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_intg_tpm3 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isTSNEdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_intg_h5_1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isTSNEdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "tsne", label = T, label.size = 3) + NoLegend()
})
}
})
output$TSNEplot_intg_h5_2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isTSNEdone_intg)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "tsne", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
observeEvent(input$doCELLiD_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running CELLiD...", value = 0.3, {
ref = readRDS(url("https://www.immunesinglecell.org/api/vishuo/download/getCellidRef"))
v$scData1.combined.rna.data.average = AverageExpression(v$scData1.combined)
v$scData1.combined.rna.data.average = round(v$scData1.combined.rna.data.average$RNA, 2)
if(input$cellatlas1a == "all"){
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, ref)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, adipose1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, adrenal_gland1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, blood1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, bone_marrow1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, brain1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, breast1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, breast_milk1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, eye1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, gut1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, heart1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, kidney1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, liver1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, lung1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, pancreas1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, PDAC1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, skin1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, testis1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, thymus1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas1a == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
v$res_intg = FastIntegration::CELLiD(v$scData1.combined.rna.data.average, tonsil1)
print(v$res_intg)
v$scData1.combined$primary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),1]
v$scData1.combined$secondary.predict = v$res_intg[as.numeric(v$scData1.combined$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res_intg) <- newheaders
print(v$scData1.combined@meta.data)
output$CELLiD2.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res_intg, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
v$isCELLiDdone1a <- TRUE
shinyalert("Celltype annotation done", "Celltype done, please perform DEG Analysis", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_cellid_intg <- renderPlotly({
if(is.null(v$isCELLiDdone1a)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", group.by = "primary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_cellid_intg1 <- renderPlotly({
if(is.null(v$isCELLiDdone1a)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", group.by = "secondary.predict", label = T, label.size = 3) + NoLegend()
})
}
})
output$ct_intg.table <- DT::renderDataTable(
v$res_intg, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
observeEvent(input$doCelltypist_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Celltypist...", value = 0.3, {
sc <- reticulate::import("scanpy", convert = FALSE)
ct <- reticulate::import("celltypist", convert = FALSE)
sceasy::convertFormat(v$scData1.combined, from = "seurat", to = "anndata", outFile = 'ct_scrna.h5ad')
v$adata = sc$read_h5ad('ct_scrna.h5ad')
v$res = ct$annotate(filename = 'ct_scrna.h5ad', model = input$celltypistatlas5, majority_voting=T)
print(v$res)
v$adata = v$res$to_adata()
print("fff")
v$adata$obs$to_csv('celltypist_predict.csv')
v$meta5 <- read.csv('celltypist_predict.csv', header = T, row.names = 1)
v$scData1.combined <- AddMetaData(v$scData1.combined, metadata = v$meta5)
v$scData1.combined$primary.predict <- v$scData1.combined$majority_voting
v$scData1.combined$secondary.predict <- v$scData1.combined$predicted_labels
print(v$scData1.combined@meta.data)
v$isCelltypistdone5 <- TRUE
shinyalert("Celltype annotation performed", "Celltype annotation performed, please perform cell-cell similarity", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_celltypist_intg <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isCelltypistdone5)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", group.by = "majority_voting", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_celltypist_intg1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isCelltypistdone5)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scData1.combined, reduction = "umap", group.by = "predicted_labels", label = T, label.size = 3) + NoLegend()
})
}
})
output$ct_celltypist_intg.table <- DT::renderDataTable(
v$meta5, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
observeEvent(input$doDeg_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
if(input$deg1a == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
ips.markers_intg <- FindAllMarkers(v$scData1.combined, only.pos = FALSE, min.pct = input$min_pct_intg, logfc.threshold = input$logfc_intg, test.use = input$test.use_intg)
v$ips.markers_intg <- ips.markers_intg
shinyalert("DEG Analysis done", "DEG Analysis done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
if(input$deg1a == "Predicted celltype"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
ips.markers_intg <- FindAllMarkers(v$scData1.combined, only.pos = FALSE, min.pct = input$min_pct_intg, logfc.threshold = input$logfc_intg, test.use = input$test.use_intg)
v$ips.markers_intg <- ips.markers_intg
shinyalert("DEG Analysis done", "DEG Analysis done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
if(input$deg1a == "Metadata celltype"){
v$scData1.combined$celltype -> Idents(v$scData1.combined)
ips.markers_intg <- FindAllMarkers(v$scData1.combined, only.pos = FALSE, min.pct = input$min_pct_intg, logfc.threshold = input$logfc_intg, test.use = input$test.use_intg)
v$ips.markers_intg <- ips.markers_intg
shinyalert("DEG Analysis done", "DEG Analysis done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
observeEvent(input$Vis_intg, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Visualizing...", value=0, {
v$isVisIntgdone <- TRUE
})
}
})
output$deg.gene.select_intg <- renderUI({
if(is.null(v$ips.markers_intg)){
return(NULL)
}else{
selectInput("deg.gene_intg", label = "Gene to visualise",
choices = unique(v$ips.markers_intg$gene), selected = rownames(v$ips.markers_intg$gene)[10])
}
})
plotViolin_intg <- reactive({
if(is.null(v$ips.markers_intg) || is.null(v$isVisIntgdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg1b == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
VlnPlot(v$scData1.combined, input$deg.gene_intg, group.by = "seurat_clusters")
}
else if(input$deg1b == "Predicted celltype"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
VlnPlot(v$scData1.combined, input$deg.gene_intg, group.by = "primary.predict")
}
else if(input$deg1b == "Metadata celltype"){
v$scData1.combined$celltype -> Idents(v$scData1.combined)
VlnPlot(v$scData1.combined, input$deg.gene_intg, group.by = "celltype")
}
})
}
})
output$Deg.plot_intg <- renderPlotly({
plotViolin_intg()
})
output$download_violn_intg <- downloadHandler(
filename = function(){"Violin plot (Integration).png"},
content = function(fname){
ggsave(fname,plotViolin_intg(), height = 7, width = 7)
}
)
plotFeature_intg <- reactive({
if(is.null(v$ips.markers_intg) || is.null(v$isVisIntgdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg1b == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
FeaturePlot(v$scData1.combined, input$deg.gene_intg)
}
else if(input$deg1b == "Predicted celltype"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
FeaturePlot(v$scData1.combined, input$deg.gene_intg)
}
else if(input$deg1b == "Metadata celltype"){
v$scData1.combined$celltype -> Idents(v$scData1.combined)
FeaturePlot(v$scData1.combined, input$deg.gene_intg)
}
})
}
})
output$Deg1.plot_intg <- renderPlotly({
plotFeature_intg()
})
output$download_feature_intg <- downloadHandler(
filename = function(){"Feature plot (Integration).png"},
content = function(fname){
ggsave(fname,plotFeature_intg(), height = 7, width = 7)
}
)
plotRidge_intg <- reactive({
if(is.null(v$ips.markers_intg) || is.null(v$isVisIntgdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg1b == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
RidgePlot(v$scData1.combined, input$deg.gene_intg, group.by = "seurat_clusters")
}
else if(input$deg1b == "Predicted celltype"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
RidgePlot(v$scData1.combined, input$deg.gene_intg, group.by = "primary.predict")
}
else if(input$deg1b == "Metadata celltype"){
v$scData1.combined$celltype -> Idents(v$scData1.combined)
RidgePlot(v$scData1.combined, input$deg.gene_intg, group.by = "celltype")
}
})
}
})
output$Deg2.plot_intg <- renderPlot({
plotRidge_intg()
})
output$download_ridge_intg <- downloadHandler(
filename = function(){"Ridge plot (Integration).png"},
content = function(fname){
ggsave(fname,plotRidge_intg(), height = 7, width = 7)
}
)
output$Deg3.plot_intg <- renderPlotly({
if(is.null(v$ips.markers_intg)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers_intg %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1.combined, features = top10$gene, size = 4) + theme(axis.text.y = element_text(size = 5)) + NoLegend()
})
}
})
output$Deg.table_intg <- DT::renderDataTable(
v$ips.markers_intg, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
}
})
output$gene1a.select <- renderUI({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
if(input$deg3a == "seurat_clusters"){
selectInput("gene1a", label = "Celltype1",
choices = as.vector(v$scData1.combined$seurat_clusters))
}
else if(input$deg3a == "Predicted celltype"){
selectInput("gene1a", label = "Celltype1",
choices = as.vector(v$scData1.combined$primary.predict))
}
else if(input$deg3a == "Metadata Celltype"){
selectInput("gene1a", label = "Celltype1",
choices = as.vector(v$scData1.combined$celltype))
}
}
})
output$gene2a.select <- renderUI({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
if(input$deg3a == "seurat_clusters"){
selectInput("gene2a", label = "Celltype2",
choices = as.vector(v$scData1.combined$seurat_clusters))
}
else if(input$deg3a == "Predicted celltype"){
selectInput("gene2a", label = "Celltype2",
choices = as.vector(v$scData1.combined$primary.predict))
}
else if(input$deg3a == "Metadata Celltype"){
selectInput("gene2a", label = "Celltype2",
choices = as.vector(v$scData1.combined$celltype))
}
}
})
observeEvent(input$doVolcano1, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
if (input$deg3a == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined_subset <- subset(v$scData1.combined, subset = seurat_clusters == input$gene1a | seurat_clusters == input$gene2a)
ips.markers_a1 <- FindAllMarkers(v$scData1.combined_subset, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "RNA", test.use = input$test.use_a)
ips.markers_b1 <- FindMarkers(v$scData1.combined_subset, ident.1 = input$gene1a, ident.2 = input$gene2a, only.pos = F, min.pct = input$min_pct_a1, logfc.threshold = input$logfc_a1, test.use = input$test.use_a1)
v$ips.markers_a1 <- ips.markers_a1
v$ips.markers_b1 <- ips.markers_b1
}
if (input$deg3a == "Predicted celltype"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
v$scData1.combined_subset <- subset(v$scData1.combined, subset = primary.predict == input$gene1a | primary.predict == input$gene2a)
ips.markers_a1 <- FindAllMarkers(v$scData1.combined_subset, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "RNA", test.use = input$test.use_a)
ips.markers_b1 <- FindMarkers(v$scData1.combined_subset, ident.1 = input$gene1a, ident.2 = input$gene2a, only.pos = F, min.pct = input$min_pct_a1, logfc.threshold = input$logfc_a1, test.use = input$test.use_a1)
v$ips.markers_a1 <- ips.markers_a1
v$ips.markers_b1 <- ips.markers_b1
}
if (input$deg3a == "Metadata celltype"){
v$scData1.combined$celltype -> Idents(v$scData1.combined)
v$scData1.combined_subset <- subset(v$scData1.combined, subset = celltype == input$gene1a | celltype == input$gene2a)
ips.markers_a1 <- FindAllMarkers(v$scData1.combined_subset, only.pos = F, min.pct = input$min_pct_a, logfc.threshold = input$logfc_a, assay = "RNA", test.use = input$test.use_a)
ips.markers_b1 <- FindMarkers(v$scData1.combined_subset, ident.1 = input$gene1a, ident.2 = input$gene2a, only.pos = F, min.pct = input$min_pct_a1, logfc.threshold = input$logfc_a1, test.use = input$test.use_a1)
v$ips.markers_a1 <- ips.markers_a1
v$ips.markers_b1 <- ips.markers_b1
}
shinyalert("Pairwise DEGs done", "Pairwise DEGs done, please run GSEA Analysis", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$volcano.plot1 <- renderPlot({
if(is.null(v$ips.markers_b1)){
return(NULL)
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
EnhancedVolcano(toptable = v$ips.markers_b1, lab = row.names(v$ips.markers_b1), x ="avg_log2FC", y ="p_val_adj", pointSize = 1, labSize = 5, legendLabSize = 12, axisLabSize = 12)
})
}
})
output$dega1.plot <- renderPlot({
if(is.null(v$ips.markers_a1)){
return(NULL)
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
v$ips.markers_a1 %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1.combined_subset, features = top10$gene, size = 5, angle = 45) + theme(axis.text.y = element_text(size = 8)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
output$gsea.ct1a.select <- renderUI({
if(is.null(v$ips.markers_intg)){
return(NULL)
}else{
selectInput("gsea.ct1a", label = "Celltype1",
choices = as.vector(v$scData1.combined$primary.predict))
}
})
output$gsea.ct2a.select <- renderUI({
if(is.null(v$ips.markers_intg)){
return(NULL)
}else{
selectInput("gsea.ct2a", label = "Celltype2",
choices = as.vector(v$scData1.combined$primary.predict))
}
})
observeEvent(input$gsea1, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating gene set enrichment analysis...", value=0, {
if(input$species_gsea1 == "Homo sapiens" & input$category_gsea1 == "H"){
v$msigdbr_hs_go1 <- msigdbr(species = "Homo sapiens", category = "H")
print(v$msigdbr_hs_go1)
v$pathways1 <- split(x = v$msigdbr_hs_go1$gene_symbol, f = v$msigdbr_hs_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Mus musculus" & input$category_gsea1 == "H"){
v$msigdbr_mm_go1 <- msigdbr(species = "Mus musculus", category = "H")
print(v$msigdbr_mm_go1)
v$pathways1 <- split(x = v$msigdbr_mm_go1$gene_symbol, f = v$msigdbr_mm_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Homo sapiens" & input$category_gsea1 == "C2"){
v$msigdbr_hs_go1 <- msigdbr(species = "Homo sapiens", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_hs_go1)
v$pathways1 <- split(x = v$msigdbr_hs_go1$gene_symbol, f = v$msigdbr_hs_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Mus musculus" & input$category_gsea1 == "C2"){
v$msigdbr_mm_go1 <- msigdbr(species = "Mus musculus", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_mm_go1)
v$pathways1 <- split(x = v$msigdbr_mm_go1$gene_symbol, f = v$msigdbr_mm_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Homo sapiens" & input$category_gsea1 == "C5"){
v$msigdbr_hs_go1 <- msigdbr(species = "Homo sapiens", category = "C5")
print(v$msigdbr_hs_go1)
v$pathways1 <- split(x = v$msigdbr_hs_go1$gene_symbol, f = v$msigdbr_hs_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Mus musculus" & input$category_gsea1 == "C5"){
v$msigdbr_mm_go1 <- msigdbr(species = "Mus musculus", category = "C5")
print(v$msigdbr_mm_go1)
v$pathways1 <- split(x = v$msigdbr_mm_go1$gene_symbol, f = v$msigdbr_mm_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Homo sapiens" & input$category_gsea1 == "C7"){
v$msigdbr_hs_go1 <- msigdbr(species = "Homo sapiens", category = "C7")
print(v$msigdbr_hs_go1)
v$pathways1 <- split(x = v$msigdbr_hs_go1$gene_symbol, f = v$msigdbr_hs_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Mus musculus" & input$category_gsea1 == "C7"){
v$msigdbr_mm_go1 <- msigdbr(species = "Mus musculus", category = "C7")
print(v$msigdbr_mm_go1)
v$pathways1 <- split(x = v$msigdbr_mm_go1$gene_symbol, f = v$msigdbr_mm_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Homo sapiens" & input$category_gsea1 == "C8"){
v$msigdbr_hs_go1 <- msigdbr(species = "Homo sapiens", category = "C8")
print(v$msigdbr_hs_go1)
v$pathways1 <- split(x = v$msigdbr_hs_go1$gene_symbol, f = v$msigdbr_hs_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
if(input$species_gsea1 == "Mus musculus" & input$category_gsea1 == "C8"){
v$msigdbr_mm_go1 <- msigdbr(species = "Mus musculus", category = "C8")
print(v$msigdbr_mm_go1)
v$pathways1 <- split(x = v$msigdbr_mm_go1$gene_symbol, f = v$msigdbr_mm_go1$gs_name)
print(v$pathways1)
v$markers1 <- FindMarkers(v$scData1.combined, ident.1 = input$gsea.ct1a, ident.2 = input$gsea.ct2a, min.pct = input$min_pct1a, logfc.threshold = input$logfc1a, test.use = input$test.use1a)
v$markers1 <- v$markers1 %>% arrange(desc(avg_log2FC))
print(v$markers1)
v$markers1.log2FC <- v$markers1$avg_log2FC
names(v$markers1.log2FC) <- row.names(v$markers1)
v$markers1.log2FC <- sort(na.omit(v$markers1.log2FC), decreasing = TRUE)
print(v$markers1.log2FC)
v$fgseaRes1 <- fgsea(pathways = v$pathways1, stats = v$markers1.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes1)
v$topPathwaysUp1 <- v$fgseaRes1[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown1 <- v$fgseaRes1[ES < 0][head(order(pval), n=10), pathway]
v$topPathways1 <- c(v$topPathwaysUp1, rev(v$topPathwaysDown1))
}
output$gsea.done <- renderText(paste0("Gene set enrichment done!"))
v$isGSEAdone1 <- TRUE
})
}
})
output$gsea1.select <- renderUI({
if(is.null(v$pathways1)){
return(NULL)
}else{
selectInput("gsea.pathway1", label = "Gene set to visualise",
choices = names(v$pathways1))
}
})
output$gsea_plot1a <- renderPlot({
if(is.null(v$pathways1) || is.null(v$isGSEAdone1)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotEnrichment(v$pathways1[[input$gsea.pathway1]], v$markers1.log2FC) + labs(title=input$gsea.pathway1)
})
}
})
output$gsea_plot1b <- renderPlot({
if(is.null(v$pathways1) || is.null(v$isGSEAdone1)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotGseaTable(v$pathways1[v$topPathways1], v$markers1.log2FC, v$fgseaRes1, gseaParam=0.5)
})
}
})
output$gsea.table1 <- DT::renderDataTable(
v$fgseaRes1, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_gsea.table1 <- downloadHandler(
filename = function(){"GSEA Results.csv"},
content = function(fname){
withProgress(message="Downloading GSEA Results...", value=0, {
fwrite(v$fgseaRes1, fname)
})
}
)
observeEvent(input$doCC1, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running cell-cell communication...", value = 0.3, {
if(input$cc1b == "primary.predict"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
v$cellphone1 <- liana_wrap(v$scData1.combined, method = input$cc_method1, resource = input$cc_resource1, assay = "RNA")
}
if(input$cc1b == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$cellphone1 <- liana_wrap(v$scData1.combined, method = input$cc_method1, resource = input$cc_resource1, assay = "RNA")
}
output$cc1.done <- renderText(paste0("Cell-cell communication done!"))
v$isCC1done <- TRUE
shinyalert("Cell-cell communication done", "Cell-cell communication done", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$CC.gene.select1 <- renderUI({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
selectInput("CC.gene1a", label = "Source",
choices = unique(v$scData1.combined$primary.predict), selected = unique(v$scData1.combined$primary.predict)[1], multiple = T, selectize = T)
}
})
output$CC.gene1.select1 <- renderUI({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
selectInput("CC.gene2a", label = "Receptor",
choices = unique(v$scData1.combined$primary.predict), selected = unique(v$scData1.combined$primary.predict)[2], multiple = T, selectize = T)
}
})
output$CC_plot1a <- renderPlot({
if(is.null(v$isCC1done)){
return(NULL)
}else{
withProgress(message="Generating cell-cell communication plot...", value=0, {
heat_freq(v$cellphone1, pallette = c("blue", "white", "red"))
#heatmaps_plot(meta_file = 'metadata.txt', pvalues_file = 'out/pvalues.txt', count_filename = 'heatmap_count.pdf', log_filename = 'heatmap_log_count.pdf', count_network_filename = 'count_network.txt', interaction_count_filename = 'interactions_count.txt', count_network_separator = '\t', interaction_count_separator = '\t', show_rownames = T, show_colnames = T, scale="none", cluster_cols = T,border_color='white', cluster_rows = T, fontsize_row=11, fontsize_col = 11, main = '',treeheight_row=0, family='Arial', treeheight_col = 0, meta_sep='\t', pvalues_sep='\t', pvalue=input$cc_pval1)
})
}
})
output$CC_plot2a <- renderPlot({
if(is.null(v$isCC1done)){
return(NULL)
}else{
withProgress(message="Generating cell-cell communication plot...", value=0, {
chord_freq(v$cellphone1, source_groups = input$CC.gene1a, target_groups = input$CC.gene2a)
})
}
})
output$cc.table1a <- DT::renderDataTable(
v$cellphone1, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
#observeEvent(input$loadexample_atacseq1, {
# withProgress(message="Loading example Data...", value=0.5, {
# tpmFiles_example_atac1 <- Read10X_h5(filename = "atac_pbmc_5k_v1/filtered_peak_bc_matrix.h5")
# meta_example_atac1 <- read.csv('atac_pbmc_5k_v1/singlecell.csv', header = T, row.names = 1)
# chrom_assay_example_atac1 <- CreateChromatinAssay(counts = tpmFiles_example_atac1, sep = c(":", "-"), genome = 'hg19', fragments = 'atac_pbmc_5k_v1/fragments.tsv.gz', min.cells = 10, min.features = 200)
# sObj_example_atac1 <- CreateSeuratObject(chrom_assay_example_atac1, assay = "peaks", project = input$projName4, meta.data = meta_example_atac1)
# print(sObj_example_atac1)
# annotations_atac <- GetGRangesFromEnsDb(ensdb = EnsDb.Hsapiens.v86)
# seqlevelsStyle(annotations_atac) <- 'UCSC'
# genome(annotations_atac) <- "hg19"
# Annotation(sObj_example_atac1) <- annotations_atac
# sObj_example_atac1 <- NucleosomeSignal(object = sObj_example_atac1)
# sObj_example_atac1 <- TSSEnrichment(object = sObj_example_atac1, fast = FALSE)
# sObj_example_atac1$pct_reads_in_peaks <- sObj_example_atac1$peak_region_fragments / sObj_example_atac1$passed_filters * 100
# sObj_example_atac1$blacklist_ratio <- sObj_example_atac1$blacklist_region_fragments / sObj_example_atac1$peak_region_fragments
# sObj_example_atac1$high.tss <- ifelse(sObj_example_atac1$TSS.enrichment > 2, 'High', 'Low')
# sObj_example_atac1$nucleosome_group <- ifelse(sObj_example_atac1$nucleosome_signal > 4, 'NS > 4', 'NS < 4')
# sObj_example_atac1 <- RunTFIDF(sObj_example_atac1)
# sObj_example_atac1 <- FindTopFeatures(sObj_example_atac1, min.cutoff = 'q0')
# sObj_example_atac1 <- RunSVD(sObj_example_atac1)
# sObj_example_atac1 <- FindNeighbors(object = sObj_example_atac1, reduction = 'lsi', dims = 2:30)
# sObj_example_atac1 <- FindClusters(object = sObj_example_atac1, resolution = 0.5, verbose = FALSE, algorithm = 3)
# sObj_example_atac1 <- RunUMAP(object = sObj_example_atac1, reduction = 'lsi', dims = 2:30)
# gene.activities1 <- GeneActivity(sObj_example_atac1, features = VariableFeatures(v$scData1.combined))
# sObj_example_atac1[["ACTIVITY"]] <- CreateAssayObject(counts = gene.activities1)
# DefaultAssay(sObj_example_atac1) <- "ACTIVITY"
# sObj_example_atac1 <- NormalizeData(sObj_example_atac1)
# sObj_example_atac1 <- ScaleData(sObj_example_atac1, features = rownames(sObj_example_atac1))
# transfer.anchors1 <- FindTransferAnchors(reference = v$scData1.combined, query = sObj_example_atac1, features = VariableFeatures(object = v$scData1.combined), reference.assay = "integrated", query.assay = "ACTIVITY", reduction = "cca")
# celltype.predictions1 <- TransferData(anchorset = transfer.anchors1, refdata = v$scData1.combined$primary.predict, weight.reduction = sObj_example_atac1[["lsi"]], dims = 2:30)
# sObj_example_atac1 <- AddMetaData(sObj_example_atac1, metadata = celltype.predictions1)
# v$transfer.anchors1 <- transfer.anchors1
# v$atacData1 <- sObj_example_atac1
# })
# label1 <- "Example loaded"
# updateActionButton(inputId = "loadexample_atacseq", label = label1)
# shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
#})
#output$annotate_scRNA_ATAC_plot_a <- renderPlot({
# if(is.null(v$scData1.combined)|| is.null(v$atacData1)){
# return(NULL)
# }else{
# withProgress(message="Generating ATAC UMAP...", value=0, {
# DimPlot(v$atacData1, group.by = "seurat_clusters", label = T)
# })
# }
#})
#output$annotate_scRNA_ATAC_plot_1a <- renderPlot({
# if(is.null(v$scData1.combined)|| is.null(v$atacData1)){
# return(NULL)
# }else{
# withProgress(message="Generating ATAC UMAP...", value=0, {
# DimPlot(v$atacData1, group.by = "predicted.id", label = T)
# })
# }
#})
#observe({
# if(input$process_atacseq1 > 0){
# print('2')
# session$sendCustomMessage("myCallbackHandler1a", "2")
# }
#})
observeEvent(input$annotate1_atacseq, {
tpmFiles_atac <- input$tpmFiles_atac
if (is.null(tpmFiles_atac)){
v$atacData1 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Annotating...", value=0.5, {
transfer.anchors1 <- FindTransferAnchors(reference = v$scData1.combined, query = v$atacData, reduction = "cca")
celltype.predictions1 <- TransferData(anchorset = transfer.anchors1, refdata = v$scData1.combined$primary.predict, weight.reduction = v$atacData1[["lsi"]], dims = 2:30)
v$atacData1 <- AddMetaData(v$atacData1, metadata = celltype.predictions1)
v$transfer.anchors1 <- transfer.anchors1
})
label1 <- "Annotation done"
updateActionButton(inputId = "loadexample_atacseq1", label = label1)
shinyalert("Annotation done", "Annotation done.", type = "success", imageWidth = 10, imageHeight = 10)
}
})
observeEvent(input$doIntg_harmony, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Data Integration...", value = 0.3, {
v$scData1.list <- SplitObject(v$scData2, split.by = "batch")
print(v$scData2)
print(v$scData1.list)
v$scData1.list <- pbmclapply(mc.cores = 20, X = v$scData1.list, FUN = function(x) {
x <- NormalizeData(x)
x <- FindVariableFeatures(x, selection.method = "vst", nfeatures = 2000)
})
features <- SelectIntegrationFeatures(object.list = v$scData1.list, nfeatures = input$nfeatures_intg_harmony)
print(features)
v$scData1.anchors <- FindIntegrationAnchors(object.list = v$scData1.list, anchor.features = features)
v$scData1.combined <- IntegrateData(anchorset = v$scData1.anchors)
print(v$scData1.anchors)
print(v$scData1.combined)
#v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "harmony", dims = 1:input$dim.used_intg)
#v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
#print(v$scData1.combined)
v$isIntgHarmonydone <- TRUE
shinyalert("Data integration done", "Data integration done, please run PCA", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$runPCA_intg_harmony, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0.3, {
DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- ScaleData(v$scData1.combined, verbose = FALSE)
v$scData1.combined <- RunPCA(v$scData1.combined, verbose = FALSE)
print(v$scData1.combined[["pca"]], dims = 1:5, nfeatures = 5)
v$scData1.combined <- RunHarmony(v$scData1.combined, "batch", assay.use = "integrated")
print(v$scData1.combined[["harmony"]], dims = 1:5, nfeatures = 5)
#output$pca1.done <- renderText(paste0("PCA done!"))
v$isPCAdone_intg_harmony <- TRUE
shinyalert("PCA done", "PCA done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$PCAplot_harmony_tpm1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_harmony)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "harmony", label = T) + NoLegend()
})
}
})
output$PCAplot_harmony_tpm2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_harmony)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "harmony", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$PCAplot_harmony_tpm3 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_harmony)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "harmony", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$PCAplot_harmony_h5_1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_harmony)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "harmony", label = T, label.size = 3) + NoLegend()
})
}
})
output$PCAplot_harmony_h5_2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_harmony)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "harmony", label = T, group.by = 'batch', label.size = 3) + NoLegend()
})
}
})
output$vizPlot_intg_harmony <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
VizDimLoadings(v$scData1.combined, dims = as.numeric(input$select.pc_intg_harmony), reduction = "harmony")
}
})
output$PCHeatmap_intg_harmony <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
DimHeatmap(v$scData1.combined, dims = as.numeric(input$select.pc_intg_harmony), reduction = "harmony")
}
})
output$PCtable_intg_harmony <- DT::renderDataTable({
if(is.null(v$scData1.combined) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_intg_harmony <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scData1.combined, reduction = "harmony", ndims = 50)
})
}
}, height = 400, width = 450)
observeEvent(input$doDeg_intg_harmony, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
v$scData1.combined$celltype -> Idents(v$scData1.combined)
ips.markers1 <- FindAllMarkers(v$scData1.combined, only.pos = FALSE, min.pct = input$min_pct_intg_harmony, logfc.threshold = input$logfc_intg_harmony, test.use = input$test.use_intg_harmony)
v$ips.markers1 <- ips.markers1
shinyalert("DEGs estimated", "DEGs estimated, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$deg.gene.select_intg_harmony <- renderUI({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
selectInput("deg.gene_intg_harmony", label = "Gene to visualise",
choices = unique(v$ips.markers1$gene))
}
})
output$Deg.plot_intg_harmony <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
VlnPlot(v$scData1.combined, input$deg.gene_intg_harmony)
})
}
})
output$Deg1.plot_intg_harmony <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(v$scData1.combined, input$deg.gene_intg_harmony)
})
}
})
output$Deg2.plot_intg_harmony <- renderPlot({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
RidgePlot(v$scData1.combined, features = input$deg.gene_intg_harmony)
})
}
})
output$Deg3.plot_intg_harmony <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers1 %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1.combined, features = top10$gene, size = 4) + theme(axis.text.y = element_text(size = 5)) + NoLegend()
})
}
})
output$Deg.table_intg_harmony <- DT::renderDataTable(
v$ips.markers1, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
#observe({if(input$scAnalysis_integ == "scVI"){
observeEvent(input$doIntg_scvi, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Data Integration...", value = 0.3, {
sc <- reticulate::import("scanpy", convert = FALSE)
scvi <- reticulate::import("scvi", convert = FALSE)
v$adata <- convertFormat(v$scData2, from="seurat", to="anndata", main_layer="counts", drop_single_values=FALSE)
print(v$adata)
scvi$model$SCVI$setup_anndata(v$adata, batch_key = 'batch')
# create the model
model = scvi$model$SCVI(v$adata, n_latent = input$scvi_latent)
print(model)
# train the model
model$train()
print(model)
v$latent = model$get_latent_representation()
print(v$latent)
# put it back in our original Seurat object
v$latent <- as.matrix(v$latent)
print(v$latent)
rownames(v$latent) = colnames(v$scData2)
#v$scData2[["scvi"]] <- CreateDimReducObject(embeddings = v$latent, key = "scvi_", assay = DefaultAssay(v$scData2))
#print(v$scData2)
v$isIntgscVIdone <- TRUE
shinyalert("Data integration done", "Data integration done, please run PCA", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$runPCA_intg_scvi, {
#tpmFiles1 <- v$scData1.combined
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0.3, {
v$scData2[["scvi"]] <- CreateDimReducObject(embeddings = v$latent, key = "scvi_", assay = DefaultAssay(v$scData2))
print(v$scData2)
v$scData2 = v$scData1.combined
print(v$scData1.combined)
#output$pca1.done <- renderText(paste0("PCA done!"))
v$isPCAdone_intg_scvi <- TRUE
shinyalert("PCA done", "PCA done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$PCAplot_scvi_tpm1 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isPCAdone_intg_scvi)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "scvi", label = T)
})
}
})
output$PCAplot_scvi_tpm2 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isPCAdone_intg_scvi)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "scvi", label = T, group.by = 'batch', label.size = 3)
})
}
})
output$PCAplot_scvi_tpm3 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isPCAdone_intg_scvi)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "scvi", label = T, group.by = 'celltype', label.size = 3)
})
}
})
output$PCAplot_scvi_h5_1 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isPCAdone_intg_scvi)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "scvi", label = T, label.size = 3)
})
}
})
output$PCAplot_scvi_h5_2 <- renderPlotly({
if(is.null(v$scData2) || is.null(v$isPCAdone_intg_scvi)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "scvi", label = T, group.by = 'batch', label.size = 3)
})
}
})
output$vizPlot_intg_scvi <- renderPlot({
if(is.null(v$scData2)){
return(NULL)
}else{
VizDimLoadings(v$scData1.combined, dims = as.numeric(input$select.pc_intg_scvi), reduction = "scvi")
}
})
output$PCHeatmap_intg_scvi <- renderPlot({
if(is.null(v$scData2)){
return(NULL)
}else{
DimHeatmap(v$scData1.combined, dims = as.numeric(input$select.pc_intg_scvi), reduction = "scvi")
}
})
output$PCtable_intg_scvi <- DT::renderDataTable({
if(is.null(v$scData2) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
observeEvent(input$doDeg_intg_scvi, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
ips.markers1 <- FindAllMarkers(v$scData1.combined, only.pos = FALSE, min.pct = input$min_pct_intg_scvi, logfc.threshold = input$logfc_intg_scvi, test.use = input$test.use_intg_scvi)
v$ips.markers1 <- ips.markers1
shinyalert("DEGs estimated", "DEGs estimated, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$deg.gene.select_intg_scvi <- renderUI({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
selectInput("deg.gene_intg_scvi", label = "Gene to visualise",
choices = unique(v$ips.markers1$gene))
}
})
output$Deg.plot_intg_scvi <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
VlnPlot(v$scData1.combined, input$deg.gene_intg_scvi)
})
}
})
output$Deg1.plot_intg_scvi <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(v$scData1.combined, input$deg.gene_intg_scvi)
})
}
})
output$Deg2.plot_intg_scvi <- renderPlot({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
RidgePlot(v$scData1.combined, features = input$deg.gene_intg_scvi)
})
}
})
output$Deg3.plot_intg_scvi <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers1 %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1.combined, features = top10$gene, size = 4) + theme(axis.text.y = element_text(size = 5)) + NoLegend()
})
}
})
output$Deg.table_intg_scvi <- DT::renderDataTable(
v$ips.markers1, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
observeEvent(input$doIntg_fastmnn, {
tpmFiles1 <- v$scData2
if (is.null(tpmFiles1)){
v$scData2 <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Data Integration...", value = 0.3, {
v$scData1.list <- SplitObject(v$scData2, split.by = "batch")
v$scData1.list <- pbmclapply(mc.cores = 20, X = v$scData1.list, FUN = function(x) {
x <- NormalizeData(x)
x <- FindVariableFeatures(x, selection.method = "vst", nfeatures = 2000)
})
v$features <- SelectIntegrationFeatures(object.list = v$scData1.list, nfeatures = input$nfeatures_intg_fastmnn)
print(v$features)
print(v$scData2)
print(v$scData1.list)
#v$scData1.combined <- RunFastMNN(object.list = v$scData1.list, features = features)
#v$scData1.combined <- FindNeighbors(v$scData1.combined, reduction = "mnn", dims = 1:input$dim.used_intg)
#v$scData1.combined <- FindClusters(v$scData1.combined, resolution = input$clus.res_intg)
#print(v$scData1.combined)
v$isIntgFastMNNdone <- TRUE
shinyalert("Data integration done", "Data integration done, please run PCA", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$runPCA_intg_fastmnn, {
tpmFiles1 <- v$scData1.list
if (is.null(tpmFiles1)){
v$scData1.list <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
#DefaultAssay(v$scData1.combined) <- "integrated"
v$scData1.combined <- RunFastMNN(object.list = v$scData1.list, features = v$features)
#output$pca1.done <- renderText(paste0("PCA done!"))
v$isPCAdone_intg_fastmnn <- TRUE
shinyalert("PCA done", "PCA done, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$PCAplot_fastmnn_tpm1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_fastmnn)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "mnn", label = T)
})
}
})
output$PCAplot_fastmnn_tpm2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_fastmnn)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "mnn", label = T, group.by = 'batch', label.size = 3)
})
}
})
output$PCAplot_fastmnn_tpm3 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_fastmnn)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "mnn", label = T, group.by = 'celltype', label.size = 3) + NoLegend()
})
}
})
output$PCAplot_fastmnn_h5_1 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_fastmnn)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "mnn", label = T, label.size = 3)
})
}
})
output$PCAplot_fastmnn_h5_2 <- renderPlotly({
if(is.null(v$scData1.combined) || is.null(v$isPCAdone_intg_fastmnn)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of integrated dataset...", value=0, {
DimPlot(v$scData1.combined, reduction = "mnn", label = T, group.by = 'batch', label.size = 3)
})
}
})
output$vizPlot_intg_fastmnn <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
VizDimLoadings(v$scData1.combined, dims = as.numeric(input$select.pc_intg_fastmnn), reduction = "mnn")
}
})
output$PCtable_intg_fastmnn <- DT::renderDataTable({
if(is.null(v$scData1.combined) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_intg_fastmnn <- renderPlot({
if(is.null(v$scData1.combined)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scData1.combined, reduction = "mnn", ndims = 50)
})
}
}, height = 400, width = 450)
observeEvent(input$doDeg_intg_fastmnn, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
v$scData1.combined$celltype -> Idents(v$scData1.combined)
ips.markers1 <- FindAllMarkers(v$scData1.combined, only.pos = FALSE, min.pct = input$min_pct_intg_fastmnn, logfc.threshold = input$logfc_intg_fastmnn, test.use = input$test.use_intg_fastmnn)
v$ips.markers1 <- ips.markers1
shinyalert("DEGs estimated", "DEGs estimated, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$deg.gene.select_intg_fastmnn <- renderUI({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
selectInput("deg.gene_intg_fastmnn", label = "Gene to visualise",
choices = unique(v$ips.markers1$gene))
}
})
output$Deg.plot_intg_fastmnn <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
VlnPlot(v$scData1.combined, input$deg.gene_intg_fastmnn)
})
}
})
output$Deg1.plot_intg_fastmnn <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(v$scData1.combined, input$deg.gene_intg_fastmnn)
})
}
})
output$Deg2.plot_intg_fastmnn <- renderPlot({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
RidgePlot(v$scData1.combined, features = input$deg.gene_intg_fastmnn)
})
}
})
output$Deg3.plot_intg_fastmnn <- renderPlotly({
if(is.null(v$ips.markers1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers1 %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData1.combined, features = top10$gene, size = 4) + theme(axis.text.y = element_text(size = 5)) + NoLegend()
})
}
})
output$Deg.table_intg_fastmnn <- DT::renderDataTable(
v$ips.markers1, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
##---------------Cell-cell similarity of Data Integration-------------------
observeEvent(input$cell_cell1, {
tpmFiles1 <- v$scData1.combined
if (is.null(tpmFiles1)){
v$scData1.combined <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating heatmap...", value=0, {
if(input$cell1a == "primary.predict"){
v$scData1.combined$primary.predict -> Idents(v$scData1.combined)
v$scData1.combined.rna.data.average1 = AverageExpression(v$scData1.combined)
v$scData1.combined.rna.data.average1 = data.frame(v$scData1.combined.rna.data.average1$RNA)
print(v$scData1.combined.rna.data.average1)
v$cor1 <- cor(v$scData1.combined.rna.data.average1, method = input$corr_method1)
print(v$cor1)
output$CELL1.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLdone1 <- TRUE
}
else if(input$cell1a == "seurat_clusters"){
v$scData1.combined$seurat_clusters -> Idents(v$scData1.combined)
v$scData1.combined.rna.data.average1 = AverageExpression(v$scData1.combined)
v$scData1.combined.rna.data.average1 = data.frame(v$scData1.combined.rna.data.average1$RNA)
print(v$scData1.combined.rna.data.average1)
v$cor1 <- cor(v$scData1.combined.rna.data.average1, method = input$corr_method1)
rownames(v$cor1) <- substr(rownames(v$cor1),2,nchar(rownames(v$cor1)))
colnames(v$cor1) <- substr(colnames(v$cor1),2,nchar(colnames(v$cor1)))
print(v$cor1)
output$CELL1.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLdone1 <- TRUE
}
if(input$cell1a == "celltype"){
v$scData1.combined$celltype -> Idents(v$scData1.combined)
v$scData1.combined.rna.data.average1 = AverageExpression(v$scData1.combined)
v$scData1.combined.rna.data.average1 = data.frame(v$scData1.combined.rna.data.average1$RNA)
print(v$scData1.combined.rna.data.average1)
v$cor1 <- cor(v$scData1.combined.rna.data.average1, method = input$corr_method1)
print(v$cor1)
output$CELL1.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLdone1 <- TRUE
}
})
}
})
plotCELL1 <- reactive({
if(is.null(v$scData1.combined.rna.data.average1) || is.null(v$isCELLdone1)){
plotly_empty()
}else{
withProgress(message="Generating Celltype similarity plot...", value=0, {
p <- heatmaply(as.matrix(v$cor1), cexRow = 0.8, cexCol = 0.8, margins = c(10,10), k_col =2, k_row = 2, colors = rev(RColorBrewer::brewer.pal(9, "RdBu")))
})
}
})
output$cell_cell_sim1 <- renderPlotly({
plotCELL1()
})
output$download_cell_cell_sim <- downloadHandler(
filename = "Celltype similarity.png",
content = function(file) {
png(file)
heatmap(as.matrix(v$cor1), col = RColorBrewer::brewer.pal(9, "RdBu"), cexRow = 0.8, cexCol = 0.8, margins = c(10,10))
dev.off()
}
)
output$cor.table1 <- DT::renderDataTable(
v$cor1, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_cor.table1 <- downloadHandler(
filename = function(){"Celltype_similarity.csv"},
content = function(fname){
withProgress(message="Downloading celltype_similarity...", value=0, {
write.csv(v$cor1, fname)
})
}
)
##------------------------Single cell multiomics module--------------------------##
output$multiomics_image <- renderImage({
list(src = "www/multiomics_fig1.png",
width = 900,
height = 550)
}, deleteFile = FALSE)
observeEvent(input$loadexample_cite, {
withProgress(message="Loading example Data...", value=0.5, {
cite.data <- Read10X_h5('multiomics/cite/pbmc_10k_protein_v3_filtered_feature_bc_matrix.h5')
})
if (is.null(cite.data)){
v$scDatat <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(cite.data)
#label1 <- "Example loaded"
#updateActionButton(inputId = "loadexample_cite_seurat", label = label1)
#shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
cite.data.rna <- cite.data$`Gene Expression`
cite.data.ab <- cite.data$`Antibody Capture`
incProgress(0.5, "Creating Seurat Object")
sObj3 <- CreateSeuratObject(cite.data.rna, project = input$projName2, min.genes = input$min.genes2, min.cells = input$min.cells2)
sObj3[["percent.mt"]] <- PercentageFeatureSet(sObj3, pattern = "^MT-")
v$scDatat <- sObj3
v$scDatat.rna <- cite.data.rna
v$scDatat.ab <- cite.data.ab
print(sObj3@meta.data)
v$scDatab <- CreateAssayObject(counts = cite.data.ab)
v$scDatat[["ADT"]] <- v$scDatab
print(Assays(v$scDatat))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_cite_seurat", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$loadexample_multiome, {
withProgress(message="Loading example Data...", value=0.5, {
multiome.data <- Read10X_h5('multiomics/multiome/pbmc_unsorted_3k_filtered_feature_bc_matrix.h5')
})
if (is.null(multiome.data)){
v$scDatan <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(multiome.data)
multiome.rna <- multiome.data$`Gene Expression`
multiome.atac <- multiome.data$Peaks
incProgress(0.5, "Creating Seurat Object")
sObj4 <- CreateSeuratObject(multiome.rna, project = input$projName2, min.genes = input$min.genes2, min.cells = input$min.cells2)
sObj4[["percent.mt"]] <- PercentageFeatureSet(sObj4, pattern = "^MT-")
print(sObj4)
print(sObj4@meta.data)
grange.counts <- StringToGRanges(rownames(multiome.atac), sep = c(":", "-"))
grange.use <- seqnames(grange.counts) %in% standardChromosomes(grange.counts)
multiome.atac <- multiome.atac[as.vector(grange.use), ]
annotations <- GetGRangesFromEnsDb(ensdb = EnsDb.Hsapiens.v86)
genome(annotations) <- "hg19"
seqlevelsStyle(annotations) <- 'UCSC'
frag.file <- "multiomics/multiome/pbmc_unsorted_3k_atac_fragments.tsv.gz"
chrom_assay <- CreateChromatinAssay(
counts = multiome.atac,
sep = c(":", "-"),
genome = 'hg19',
fragments = frag.file,
min.cells = 10,
annotation = annotations)
print(chrom_assay)
sObj4[["ATAC"]] <- chrom_assay
print(sObj4)
v$scDatan <- sObj4
v$scDatan.rna <- multiome.rna
v$scDatan.atac <- multiome.atac
print(Assays(v$scDatan))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_multiome_seurat", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$loadButton2, {
if(input$scAnalysis_type == "CITE-seq"){
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFiles2$datapath)
print(tpmFiles2$name)
print(file.exists(paste(tpmFiles2$datapath[1], "/", tpmFiles2$name[1], sep="")))
exp.data3 <- Read10X_h5(tpmFiles2$datapath)
exp.data3.rna <- exp.data3$`Gene Expression`
exp.data3.ab <- exp.data3$`Antibody Capture`
#additional.ident <- NULL
incProgress(0.5, "Creating Seurat Object")
sObj3 <- CreateSeuratObject(exp.data3.rna, project = input$projName2, min.genes = input$min.genes2, min.cells = input$min.cells2)
#mito.genes <- grep("^MT-", rownames(sObj@assays$RNA@data), ignore.case = TRUE, value = TRUE)
sObj3[["percent.mt"]] <- PercentageFeatureSet(sObj3, pattern = "^MT-")
v$scDatat <- sObj3
v$scDatat.rna <- exp.data3.rna
v$scDatat.ab <- exp.data3.ab
print(sObj3@meta.data)
v$scDatab <- CreateAssayObject(counts = exp.data3.ab)
v$scDatat[["ADT"]] <- v$scDatab
print(Assays(v$scDatat))
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
})
output$countdataDT2a <- renderDataTable({
if(!is.null(v$scDatat.rna))
{
if(ncol(v$scDatat.rna) > 20 )
return(as.matrix(v$scDatat@assays$RNA@counts[,1:20]))
}
}, server = FALSE)
output$countdataDT2b <- renderDataTable({
if(!is.null(v$scDatat.ab))
{
if(ncol(v$scDatat.ab) > 20 )
return(as.matrix(v$scDatat@assays$ADT@counts[,1:20]))
}
}, server = FALSE)
observeEvent(input$filter_seurat2a, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scDatat)
v$scDatat <- subset(v$scDatat, subset = nFeature_RNA > input$obsa & nFeature_RNA < input$obsa1 & nCount_ADT > input$obsa3 & nCount_ADT < input$obsa4 & percent.mt < input$obsa2)
print(v$scDatat)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
)
observeEvent(input$reset_mult, {
session$reload()
print("Reset done")
})
output$nFeature_RNAPlot2a <- renderPlot({
if(is.null(v$scDatat)){
plotly_empty()
}else{
VlnPlot(v$scDatat, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3) + NoLegend()
}
})
output$nFeature_RNAPlot2c <- renderPlot({
if(is.null(v$scDatat)){
plotly_empty()
}else{
VlnPlot(v$scDatat, features = c("nFeature_ADT", "nCount_ADT"), ncol = 2) + NoLegend()
}
})
output$nFeature_RNAPlot2b <- renderPlot({
if(is.null(v$scDatan)){
plotly_empty()
}else{
VlnPlot(v$scDatan, features = c("nFeature_RNA", "nCount_RNA", "percent.mt"), ncol = 3) + NoLegend()
}
})
output$nFeature_RNAPlot2d <- renderPlot({
if(is.null(v$scDatan)){
plotly_empty()
}else{
VlnPlot(v$scDatan, features = c("nFeature_ATAC", "nCount_ATAC"), ncol = 2) + NoLegend()
}
})
#output$FeatureScatterPlot_mult_a <- renderPlotly({
# if(is.null(v$scDatat)){
# plotly_empty()
# }else{
# print(FeatureScatter(v$scDatat, "nCount_RNA", "nFeature_RNA"))
# }
#})
#output$FeatureScatterPlot_mult_b <- renderPlotly({
# if(is.null(v$scDatat)){
# plotly_empty()
# }else{
# print(FeatureScatter(v$scDatat, "nCount_RNA", "percent.mt"))
# }
#})
#output$FeatureScatterPlot_mult_c <- renderPlotly({
# if(is.null(v$scDatan)){
# plotly_empty()
# }else{
# print(FeatureScatter(v$scDatan, "nCount_RNA", "nCount_ATAC"))
# }
#})
#output$FeatureScatterPlot_mult_d <- renderPlotly({
# if(is.null(v$scDatan)){
# plotly_empty()
# }else{
# print(FeatureScatter(v$scDatan, "nCount_RNA", "percent.mt"))
# }
#})
observeEvent(input$findVarGenes_mult, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding variable genes...", value = 0, {
DefaultAssay(v$scDatat) <- 'RNA'
v$scDatat <- NormalizeData(v$scDatat)
v$scDatat <- FindVariableFeatures(v$scDatat,
mean.function = ExpMean,
dispersion.function = LogVMR,
nfeatures = input$var.genes_mult,
selection.method = input$selection.method)
#all.genes <- rownames(v$scData1)
v$scDatat <- ScaleData(v$scDatat)
incProgress(0.5)
DefaultAssay(v$scDatat) <- 'ADT'
VariableFeatures(v$scDatat) <- rownames(v$scDatat[["ADT"]])
v$scDatat <- NormalizeData(v$scDatat, normalization.method = 'CLR', margin = 2)
v$scDatat <- FindVariableFeatures(v$scDatat,
mean.function = ExpMean,
dispersion.function = LogVMR,
nfeatures = input$var.genes_mult,
selection.method = input$selection.method)
#all.genes <- rownames(v$scData1)
v$scDatat <- ScaleData(v$scDatat)
DefaultAssay(v$scDatat) <- 'RNA'
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData1@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scDatat)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scDatat), 10)
variable_feature1 <- VariableFeaturePlot(v$scDatat)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("Highly variable features identified", "Highly variable features identified, please perform PCA", type = "success", imageWidth = 10, imageHeight = 10)
#dev.off()
})
}
}
output$VarGenes_mult <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scDatat)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Multiomics_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scDatat)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scDatat@assays$RNA@var.features, file = txtfile)
})
}
})
})
}
})
observeEvent(input$runPCA_mult_seurat, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0,{
incProgress(0.5, message = "Running PCA...")
DefaultAssay(v$scDatat) <- 'RNA'
v$scDatat <- RunPCA(v$scDatat, verbose = FALSE)
print(v$scDatat[["pca"]], dims = 1:5, nfeatures = 5)
DefaultAssay(v$scDatat) <- 'ADT'
v$scDatat <- RunPCA(v$scDatat, reduction.name = 'apca')
DefaultAssay(v$scDatat) <- 'RNA'
v$isPCAdone1 <- TRUE
PCA_plot1a <- DimPlot(v$scDatat, reduction = "pca", label = T)
print(PCA_plot1a)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scDatat)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
shinyalert("PCA performed", "PCA performed, please perform clustering", type = "success", imageWidth = 10, imageHeight = 10)
}
)
}
})
output$PCAplot_mult_seurat_h5_1 <- renderPlotly({
if(is.null(v$isPCAdone1)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of multiomics dataset...", value=0, {
DimPlot(v$scDatat, reduction = "pca", label = T)
})
}
})
output$vizPlot_mult_seurat <- renderPlot({
if(is.null(v$scDatat)){
return(NULL)
}else{
VizDimLoadings(v$scDatat, dims = as.numeric(input$select.pc_mult_seurat))
}
})
output$PCHeatmap_mult_seurat <- renderPlot({
if(is.null(v$scDatat)){
return(NULL)
}else{
DimHeatmap(v$scDatat, dims = as.numeric(input$select.pc_mult_seurat))
}
})
output$PCtable_mult_seurat <- DT::renderDataTable({
if(is.null(v$scDatat) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_mult_seurat <- renderPlot({
if(is.null(v$scDatat)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scDatat, ndims = 50)
})
}
}, height = 400, width = 450)
observeEvent(input$runMOFA_mult, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0,{
incProgress(0.5, message = "Running PCA...")
mofa <- reticulate::import("mofapy2", convert = FALSE)
v$mofa <- create_mofa(v$scDatat, assays = c("RNA","ADT"))
print(v$mofa)
model_opts <- get_default_model_options(v$mofa)
model_opts$num_factors <- as.numeric(input$num_factor_mofa)
v$mofa <- prepare_mofa(v$mofa, model_options = model_opts)
print(model_opts)
v$mofa <- run_mofa(v$mofa, use_basilisk = F)
v$isMOFAdone <- TRUE
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
shinyalert("MOFA2 performed", "MOFA2 performed, please perform clustering", type = "success", imageWidth = 10, imageHeight = 10)
}
)}
})
output$mofaplot_mult_1 <- renderPlot({
if(is.null(v$isMOFAdone)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of multiomics dataset...", value=0, {
plot_data_overview(v$mofa)
})
}
})
output$mofaplot_mult_2 <- renderPlot({
if(is.null(v$isMOFAdone)){
return(NULL)
}else{
plot_factor_cor(v$mofa)
}
})
output$mofaplot_mult_3 <- renderPlot({
if(is.null(v$isMOFAdone)){
return(NULL)
}else{
plot_variance_explained(v$mofa, max_r2=as.numeric(input$num_factor_mofa))
}
})
output$mofaplot_mult_4 <- renderPlot({
if(is.null(v$isMOFAdone)){
return(NULL)
}else{
plot_factors(v$mofa, factors = c(as.numeric(input$factor1),as.numeric(input$factor2)), color_by = "Factor1")
}
})
output$mofaplot_mult_5 <- renderPlot({
if(is.null(v$isMOFAdone)){
return(NULL)
}else{
plot_data_heatmap(v$mofa,
view = input$selection.view,
factor = as.numeric(input$factor1a),
features = as.numeric(input$num.features.mofa),
denoise = TRUE,
cluster_rows = TRUE, cluster_cols = FALSE,
show_rownames = TRUE, show_colnames = FALSE,
scale = "row")
}
})
observeEvent(input$findCluster_mult_seurat, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
DefaultAssay(v$scDatat) <- "RNA"
v$scDatat <- FindNeighbors(v$scDatat, reduction = "pca", dims = 1:input$dim.used_mult_seurat)
v$scDatat <- FindClusters(v$scDatat, resolution = input$clus.res_mult_seurat)
DefaultAssay(v$scDatat) <- "ADT"
#v$scDatat <- FindMultiModalNeighbors(v$scDatat, reduction.list = list("pca", "apca"), dims.list = list(1:input$dim.used_mult_seurat, 1:input$dim.used_mult_seurat), modality.weight.name = "RNA.weight")
v$scDatat <- FindClusters(v$scDatat, graph.name = "wsnn", algorithm = 3, resolution = input$clus.res_mult_seurat2a, verbose = FALSE)
#v$scDatat <- RunUMAP(v$scDatat, nn.name = "weighted.nn", reduction.name = "wnn.umap", reduction.key = "wnnUMAP_")
#output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterMultdone <- TRUE
shinyalert("Clustering performed", "Clustering performed, please run CellType Identification", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Cluster2DPlot_mult_seurat_a <- renderPlotly({
if(is.null(v$isClusterMultdone)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scDatat, reduction = "rna.umap", label = T)
})
}
})
output$Cluster2DPlot_mult_seurat_b <- renderPlotly({
if(is.null(v$isClusterMultdone)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scDatat, reduction = "adt.umap", label = T)
})
}
})
output$Cluster2DPlot_mult_seurat_c <- renderPlotly({
if(is.null(v$isClusterMultdone)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scDatat, reduction = "wnn.umap", label = T)
})
}
})
observeEvent(input$findCluster_mofa2, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(v$isMOFAdone) || is.null(v$scDatat)){
#v$scDatat <- NULL
plotly_empty()
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
cluster <- cluster_samples(v$mofa, k=input$num.clus.mofa, factors="all")
clusters <- cluster$cluster
v$clusters <- clusters
samples_metadata(v$mofa) <- as.data.frame(v$clusters) %>% tibble::rownames_to_column("sample") %>% as.data.table
v$mofa@samples_metadata -> df
rownames(df) <- df[,1]
v$scDatat <- MOFA2::add_mofa_factors_to_seurat(mofa_object = v$mofa, seurat_object = v$scDatat, views = "all", factors = "all")
v$scDatat <- AddMetaData(v$scDatat, metadata = df)
v$scDatat$v.clusters -> Idents(v$scDatat)
print(v$scDatat@meta.data)
v$isClusterMOFAdone <- TRUE
shinyalert("Clustering performed", "Clustering performed, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Cluster2DPlot_mofa <- renderPlot({
if(is.null(v$isClusterMOFAdone) || is.null(v$scDatat)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scDatat, reduction = "MOFAUMAP", group.by = "v.clusters", label = F)
})
}
})
observeEvent(input$runUMAP_mult_seurat, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0,{
incProgress(0.5, message = "Running UMAP...")
DefaultAssay(v$scDatat) <- "RNA"
v$scDatat <- RunUMAP(v$scDatat, reduction = "pca", reduction.name = "rna.umap", dims = 1:input$dim.used_mult_seurat)
DefaultAssay(v$scDatat) <- "ADT"
#VariableFeatures(v$scDatat) <- rownames(v$scDatat[["ADT"]])
#v$scDatat <- NormalizeData(v$scDatat, normalization.method = 'CLR', margin = 2) %>% ScaleData() %>%
v$scDatat <- RunUMAP(v$scDatat, reduction = "apca", reduction.name = "adt.umap", dims = 1:input$dim.used_mult_seurat)
v$scDatat <- FindMultiModalNeighbors(v$scDatat, reduction.list = list("pca", "apca"), dims.list = list(1:input$dim.used_mult_seurat, 1:input$dim.used_mult_seurat), modality.weight.name = "RNA.weight")
#v$scDatat <- FindClusters(v$scDatat, graph.name = "wsnn", algorithm = 3, resolution = input$clus.res_mult_seurat2a, verbose = FALSE)
v$scDatat <- RunUMAP(v$scDatat, nn.name = "weighted.nn", reduction.name = "wnn.umap", reduction.key = "wnnUMAP_")
v$isUMAPdone <- TRUE
print(v$scDatat)
UMAP_plot_cite_a <- DimPlot(v$scDatat, reduction = "rna.umap", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("RNA")
UMAP_plot_cite_b <- DimPlot(v$scDatat, reduction = "adt.umap", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("ADT")
UMAP_plot_cite_c <- DimPlot(v$scDatat, reduction = "wnn.umap", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("WNN")
print(UMAP_plot_cite_a)
print(UMAP_plot_cite_b)
print(UMAP_plot_cite_c)
shinyalert("UMAP done", "UMAP done, please run Clustering", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$UMAPplot_mult_seurat_1 <- renderPlot({
if(is.null(v$isUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatat, reduction = "rna.umap", label = F)
})
}
})
output$UMAPplot_mult_seurat_2 <- renderPlot({
if(is.null(v$isUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatat, reduction = "adt.umap", label = F)
})
}
})
output$UMAPplot_mult_seurat_3 <- renderPlot({
if(is.null(v$isUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatat, reduction = "wnn.umap", label = F)
})
}
})
observeEvent(input$runUMAP_mofa2, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0.3, {
v$factors <- 1:get_dimensions(v$mofa)[["K"]]
v$mofa <- run_umap(v$mofa, factors = v$factors, n_neighbors = input$num.neighbors_mofa2, min_dist = 0.30)
v$isMOFAUMAPdone <- TRUE
shinyalert("UMAP performed", "Clustering performed, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$UMAPplot_mofa_1 <- renderPlot({
if(is.null(v$isMOFAUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot...", value=0, {
plot_dimred(v$mofa, method = "UMAP")
})
}
})
observeEvent(input$runTSNE_mult_seurat, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running TSNE...", value = 0,{
incProgress(0.5, message = "Running TSNE...")
DefaultAssay(v$scDatat) <- "RNA"
v$scDatat <- RunTSNE(v$scDatat, reduction = "pca", reduction.name = "tsne.rna", dims = 1:input$dim.used_mult_seurat)
DefaultAssay(v$scDatat) <- "ADT"
v$scDatat <- RunTSNE(v$scDatat, reduction = 'apca', dims = 1:input$dim.used_mult_seurat, reduction.name = "tsne.adt", reduction.key = "adtTSNE_")
v$scDatat <- RunTSNE(v$scDatat, dims = 1:input$dim.used_mult_seurat, reduction.name = "wnn.tsne", reduction.key = "wnnTSNE_", nn.name = "weighted.nn")
v$isTSNEdone <- TRUE
TSNE_plot_cite_a <- DimPlot(v$scDatat, reduction = "tsne.rna", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("RNA")
TSNE_plot_cite_b <- DimPlot(v$scDatat, reduction = "tsne.adt", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("ADT")
TSNE_plot_cite_c <- DimPlot(v$scDatat, reduction = "wnn.tsne", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("WNN")
print(TSNE_plot_cite_a)
print(TSNE_plot_cite_b)
print(TSNE_plot_cite_c)
shinyalert("tSNE done", "tSNE done, please run Cell type identification", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$TSNEplot_mult_seurat_1 <- renderPlotly({
if(is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatat, reduction = "tsne.rna", label = T)
})
}
})
output$TSNEplot_mult_seurat_2 <- renderPlotly({
if(is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatat, reduction = "tsne.adt", label = T)
})
}
})
output$TSNEplot_mult_seurat_3 <- renderPlotly({
if(is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatat, reduction = "wnn.tsne", label = T)
})
}
})
observeEvent(input$runTSNE_mofa2, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0.3, {
v$mofa <- run_tsne(v$mofa, factors = "all", groups = "all")
samples_metadata(v$mofa) <- as.data.frame(v$clusters) %>% tibble::rownames_to_column("sample") %>% as.data.table
v$mofa@samples_metadata -> df
rownames(df) <- df[,1]
v$t1 <- MOFA2::add_mofa_factors_to_seurat(mofa_object = v$mofa, seurat_object = v$scDatat, views = "all", factors = "all")
v$t1 <- AddMetaData(v$t1, metadata = df)
print(v$t1@meta.data)
v$isMOFAdone <- TRUE
shinyalert("UMAP performed", "Clustering performed, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$TSNEplot_mofa_1 <- renderPlotly({
if(is.null(v$isMOFAdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$t1, group.by = "v.clusters", label = T, reduction = "MOFATSNE")
})
}
})
observeEvent(input$doCELLiD_mult_cite_seurat, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running CELLiD...", value = 0.3, {
ref = readRDS(url("https://www.immunesinglecell.org/api/vishuo/download/getCellidRef"))
v$scDatat.rna.data.average = AverageExpression(v$scDatat)
v$scDatat.rna.data.average = round(v$scDatat.rna.data.average$RNA, 2)
if(input$cellatlas_mult_cite_seurat == "all"){
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, ref)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, adipose1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, adrenal_gland1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, blood1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, bone_marrow1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, brain1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, breast1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, breast_milk1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, eye1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, gut1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, heart1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, kidney1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, liver1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, lung1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, pancreas1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, PDAC1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, skin1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, testis1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, thymus1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_seurat == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, tonsil1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
v$isCELLiDMultSeuratdone <- TRUE
shinyalert("Cell type identification done", "Cell type identification done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_cellid_mult_cite_seurat <- renderPlot({
if(is.null(v$scDatat) || is.null(v$isCELLiDMultSeuratdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scDatat, reduction = input$assay_mult_cite_seurat, group.by = "primary.predict", label = F, label.size = 3)
})
}
})
output$Umap_cellid_mult_cite_seurat1 <- renderPlot({
if(is.null(v$scDatat) || is.null(v$isCELLiDMultSeuratdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scDatat, reduction = input$assay_mult_cite_seurat, group.by = "secondary.predict", label = F, label.size = 3)
})
}
})
output$ct_cite_seurat.table <- DT::renderDataTable(
v$res, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$download_cellid_cite_seurat_prediction <- downloadHandler(
filename = function(){"CELLiD predictions.csv"},
content = function(fname){
withProgress(message="Downloading CELLiD predictions...", value=0, {
write.csv(v$res, fname)
})
}
)
observeEvent(input$doCELLiD_mult_cite_mofa, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running CELLiD...", value = 0.3, {
ref = readRDS(url("https://www.immunesinglecell.org/api/vishuo/download/getCellidRef"))
v$scDatat.rna.data.average = AverageExpression(v$scDatat)
v$scDatat.rna.data.average = round(v$scDatat.rna.data.average$RNA, 2)
if(input$cellatlas_mult_cite_mofa == "all"){
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, ref)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, adipose1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, adrenal_gland1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, blood1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, bone_marrow1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, brain1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, breast1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, breast_milk1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, eye1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, gut1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, heart1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, kidney1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, liver1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, lung1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, pancreas1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, PDAC1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, skin1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, testis1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, thymus1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_cite_mofa == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
v$res = FastIntegration::CELLiD(v$scDatat.rna.data.average, tonsil1)
print(v$res)
v$scDatat$primary.predict = v$res[as.numeric(v$scDatat$v.clusters),1]
v$scDatat$secondary.predict = v$res[as.numeric(v$scDatat$v.clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatat@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
v$isCELLiDMultmofadone <- TRUE
shinyalert("Cell type identification done", "Cell type identification done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_cellid_mult_cite_mofa <- renderPlot({
if(is.null(v$scDatat) || is.null(v$isCELLiDMultmofadone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scDatat, group.by = "primary.predict", label = F, label.size = 3)
})
}
})
output$Umap_cellid_mult_cite_mofa1 <- renderPlot({
if(is.null(v$scDatat) || is.null(v$isCELLiDMultmofadone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scDatat, group.by = "secondary.predict", label = F, label.size = 3)
})
}
})
output$ct_cite_mofa.table <- DT::renderDataTable(
v$res, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$download_cellid_cite_mofa_prediction <- downloadHandler(
filename = function(){"CELLiD predictions.csv"},
content = function(fname){
withProgress(message="Downloading CELLiD predictions...", value=0, {
write.csv(v$res, fname)
})
}
)
observeEvent(input$doCelltypist_mult_cite_seurat, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running Celltypist...", value = 0.3, {
sc <- reticulate::import("scanpy", convert = FALSE)
ct <- reticulate::import("celltypist", convert = FALSE)
sceasy::convertFormat(v$scDatat, from = "seurat", to = "anndata", outFile = 'ct_scrna.h5ad')
v$adata = sc$read_h5ad('ct_scrna.h5ad')
v$res = ct$annotate(filename = 'ct_scrna.h5ad', model = input$celltypistatlas9, majority_voting=T)
print(v$res)
v$adata = v$res$to_adata()
print("fff")
v$adata$obs$to_csv('celltypist_predict.csv')
v$meta9 <- read.csv('celltypist_predict.csv', header = T, row.names = 1)
v$scDatat <- AddMetaData(v$scDatat, metadata = v$meta9)
v$scDatat$primary.predict <- v$scDatat$majority_voting
v$scDatat$secondary.predict <- v$scDatat$predicted_labels
print(v$scDatat@meta.data)
v$isCelltypistdone9 <- TRUE
shinyalert("Celltype annotation performed", "Celltype annotation performed, please perform cell-cell similarity", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_celltypist_mult_cite_seurat <- renderPlotly({
if(is.null(v$scDatat) || is.null(v$isCelltypistdone9)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scDatat, reduction = input$assay_mult1_cite_seurat, group.by = "majority_voting", label = T, label.size = 3) + NoLegend()
})
}
})
output$Umap_celltypist_mult_cite_seurat1 <- renderPlotly({
if(is.null(v$scDatat) || is.null(v$isCelltypistdone9)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from Celltypist...", value=0, {
DimPlot(v$scDatat, reduction = input$assay_mult1_cite_seurat, group.by = "predicted_labels", label = T, label.size = 3) + NoLegend()
})
}
})
output$ct_celltypist_mult_cite_seurat.table <- DT::renderDataTable(
v$meta9, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$download_celltypist_cite_seurat_prediction <- downloadHandler(
filename = function(){"Celltypist predictions.csv"},
content = function(fname){
withProgress(message="Downloading CELLiD predictions...", value=0, {
write.csv(v$meta9, fname)
})
}
)
#output$vis.gene.select1 <- renderUI({
# if(is.null(v$scDatat)){
# return(NULL)
# }else{
# selectInput("vis.gene1", label = "Gene to visualise",
# choices = rownames(v$scDatat[["ADT"]]))
# }
#})
#output$vis1.plot <- renderPlotly({
# if(is.null(v$scDatat)){
# return(NULL)
# }else{
# withProgress(message="Generating DEG Plot...", value=0, {
# FeaturePlot(v$scDatat, input$vis.gene1, cols = c("lightgrey", "darkgreen"), order = T, reduction = "adt.umap")
# })
# }
#})
#observeEvent(input$Vis3, {
# tpmFiles2 <- input$tpmFiles2
# tpmFiles2 <- v$scDatat
# if (is.null(tpmFiles2)){
# v$scDatat <- NULL
# shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
# }else{
# withProgress(message = "Visualizing...", value = 0,{
# incProgress(0.5, message = "Visualizing...")
# DefaultAssay(v$scDatat) <- "ADT"
# })
# }
#})
observeEvent(input$Vis3a, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Visualizing...", value = 0,{
incProgress(0.5, message = "Visualizing...")
DefaultAssay(v$scDatat) <- "ADT"
})
}
})
output$vis.gene.select_a <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
selectInput("vis.gene_a", label = "Gene to visualise",
choices = rownames(v$scDatat[["RNA"]]))
}
})
output$vis.plot_a <- renderPlotly({
if(is.null(v$scDatat)){
return(NULL)
}else{
withProgress(message="Generating Feature Plot...", value=0, {
FeaturePlot(v$scDatat, input$vis.gene_a, cols = c("lightgrey", "darkgreen"), order = T, reduction = "MOFAUMAP")
})
}
})
output$vis.gene.select1_a <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
selectInput("vis.gene1a", label = "Gene to visualise",
choices = rownames(v$scDatat[["ADT"]]))
}
})
output$vis1.plot_a <- renderPlotly({
if(is.null(v$scDatat)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(v$scDatat, input$vis.gene1a, cols = c("lightgrey", "darkgreen"), order = T, reduction = "MOFAUMAP")
})
}
})
observeEvent(input$cell_cell_mult, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating heatmap...", value=0, {
if(input$cell_m1 == "primary.predict"){
v$scDatat$primary.predict -> Idents(v$scDatat)
v$scDatat.rna.data.average1 = AverageExpression(v$scDatat)
v$scDatat.rna.data.average1 = data.frame(v$scDatat.rna.data.average1$RNA)
print(v$scDatat.rna.data.average1)
v$cor_mult <- cor(v$scDatat.rna.data.average1, method = input$corr_method2)
print(v$cor_mult)
output$CELL_mult.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLMultdone <- TRUE
}
if(input$cell_m1 == "seurat_clusters"){
v$scDatat$seurat_clusters -> Idents(v$scDatat)
v$scDatat.rna.data.average1 = AverageExpression(v$scDatat)
v$scDatat.rna.data.average1 = data.frame(v$scDatat.rna.data.average1$RNA)
print(v$scDatat.rna.data.average1)
v$cor_mult <- cor(v$scDatat.rna.data.average1, method = input$corr_method2)
rownames(v$cor_mult) <- substr(rownames(v$cor_mult),2,nchar(rownames(v$cor_mult)))
colnames(v$cor_mult) <- substr(colnames(v$cor_mult),2,nchar(colnames(v$cor_mult)))
print(v$cor_mult)
output$CELL_mult.done <- renderText(paste0("Celltype similarity done!"))
v$isCELLMultdone <- TRUE
}
})
}
})
plotCELLMult <- reactive({
if(is.null(v$scDatat.rna.data.average1) || is.null(v$isCELLMultdone)){
plotly_empty()
}else{
withProgress(message="Generating Celltype similarity plot...", value=0, {
p <- heatmaply(as.matrix(v$cor_mult), cexRow = 0.8, cexCol = 0.8, margins = c(10,10), k_col =2, k_row = 2, colors = rev(RColorBrewer::brewer.pal(9, "RdBu")))
})
}
})
output$cell_cell_mult_sim <- renderPlotly({
plotCELLMult()
})
output$download_cell_cell_mult_sim <- downloadHandler(
filename = "Celltype similarity.png",
content = function(file) {
png(file)
heatmap(as.matrix(v$cor_mult), col = RColorBrewer::brewer.pal(9, "RdBu"), cexRow = 0.8, cexCol = 0.8, margins = c(10,10))
dev.off()
}
)
output$cor_mult.table <- DT::renderDataTable(
v$cor_mult, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_cor_mult.table <- downloadHandler(
filename = function(){"Celltype_similarity.csv"},
content = function(fname){
withProgress(message="Downloading celltype_similarity...", value=0, {
write.csv(v$cor_mult, fname)
})
}
)
observeEvent(input$doDeg_mult, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
if(input$deg_mult1 == "seurat_clusters"){
v$scDatat$seurat_clusters -> Idents(v$scDatat)
ips.markers_mult <- FindAllMarkers(v$scDatat, only.pos = FALSE, min.pct = input$min_pct_mult, logfc.threshold = input$logfc_mult, assay = "RNA", test.use = input$test.use_mult)
v$ips.markers_mult <- ips.markers_mult
shinyalert("DEGs estimated", "DEGs estimated, please perform do data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
if(input$deg_mult1 == "primary.predict"){
v$scDatat$primary.predict -> Idents(v$scDatat)
ips.markers_mult <- FindAllMarkers(v$scDatat, only.pos = FALSE, min.pct = input$min_pct_mult, logfc.threshold = input$logfc_mult, assay = "RNA", test.use = input$test.use_mult)
v$ips.markers_mult <- ips.markers_mult
shinyalert("DEGs estimated", "DEGs estimated, please perform do data visualization", type = "success", imageWidth = 10, imageHeight = 10)
}
})
}
})
observeEvent(input$Vis_mult, {
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Visualizing...", value=0, {
v$isVisMultdone <- TRUE
})
}
})
output$vis.gene.select <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
selectInput("vis.gene", label = "Gene to visualise",
choices = rownames(v$scDatat[[input$assay_mult_cite_seurat2]]))
}
})
plotFeature_mult <- reactive({
if(is.null(v$scDatat) || is.null(v$isVisMultdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg_mult2 == "seurat_clusters"){
v$scDatat$seurat_clusters -> Idents(v$scDatat)
FeaturePlot(v$scDatat, input$vis.gene, cols = c("lightgrey", "darkgreen"), order = T, reduction = input$assay_mult_cite_seurat1)
}
else if(input$deg_mult2 == "primary.predict"){
v$scDatat$primary.predict -> Idents(v$scDatat)
FeaturePlot(v$scDatat, input$vis.gene, cols = c("lightgrey", "darkgreen"), order = T, reduction = input$assay_mult_cite_seurat1)
}
})
}
})
output$vis.plot <- renderPlotly({
plotFeature_mult()
})
output$download_feature_mult <- downloadHandler(
filename = function(){"Feature plot (Multiomics module).png"},
content = function(fname){
ggsave(fname,plotFeature_mult(), height = 7, width = 7)
}
)
plotViolin_mult <- reactive({
if(is.null(v$scDatat) || is.null(v$isVisMultdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg_mult2 == "seurat_clusters"){
v$scDatat$seurat_clusters -> Idents(v$scDatat)
print(v$scDatat)
VlnPlot(v$scDatat, input$vis.gene, assay = input$assay_mult_cite_seurat2)
}
else if(input$deg_mult2 == "primary.predict"){
v$scDatat$primary.predict -> Idents(v$scDatat)
VlnPlot(v$scDatat, input$vis.gene, assay = input$assay_mult_cite_seurat2)
}
})
}
})
output$Deg_mult.plot <- renderPlotly({
plotViolin_mult()
})
output$download_violn_mult <- downloadHandler(
filename = function(){"Violin plot (Multiomics module).png"},
content = function(fname){
ggsave(fname,plotViolin_mult(), height = 7, width = 7)
}
)
plotRidge_mult <- reactive({
if(is.null(v$scDatat) || is.null(v$isVisMultdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
if(input$deg_mult2 == "seurat_clusters"){
v$scDatat$seurat_clusters -> Idents(v$scDatat)
RidgePlot(v$scDatat, features = input$vis.gene, assay = input$assay_mult_cite_seurat2)
}
else if(input$deg_mult2 == "primary.predict"){
v$scDatat$primary.predict -> Idents(v$scDatat)
RidgePlot(v$scDatat, features = input$vis.gene, assay = input$assay_mult_cite_seurat2)
}
})
}
})
output$Deg_mult2.plot <- renderPlot({
plotRidge_mult()
})
output$download_ridge_mult <- downloadHandler(
filename = function(){"Ridge plot (Multiomics module).png"},
content = function(fname){
ggsave(fname,plotRidge_mult(), height = 7, width = 7)
}
)
output$Deg3_mult.plot <- renderPlot({
if(is.null(v$ips.markers_mult)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers_mult %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scDatat, features = top10$gene, assay = "RNA", size = 5, angle = 45) + theme(axis.text.y = element_text(size = 4)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
output$Deg_mult.table <- DT::renderDataTable(
v$ips.markers_mult, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$gene1_mult.select <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
if(input$deg3_mult == "seurat_clusters"){
selectInput("gene1_mult", label = "Celltype1",
choices = as.vector(v$scDatat$seurat_clusters))
}
else if(input$deg3_mult == "primary.predict"){
selectInput("gene1_mult", label = "Celltype1",
choices = as.vector(v$scDatat$primary.predict))
}
}
})
output$gene2_mult.select <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
if(input$deg3_mult == "seurat_clusters"){
selectInput("gene2_mult", label = "Celltype2",
choices = as.vector(v$scDatat$seurat_clusters))
}
else if(input$deg3_mult == "primary.predict"){
selectInput("gene2_mult", label = "Celltype2",
choices = as.vector(v$scDatat$primary.predict))
}
}
})
observeEvent(input$doVolcano_mult, {
tpmFiles2 <- input$tpmFiles2
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
v$scDatat_subset <- subset(v$scDatat, subset = primary.predict == input$gene1_mult | primary.predict == input$gene2_mult)
ips.markers_mult_a <- FindAllMarkers(v$scDatat_subset, only.pos = F, min.pct = input$min_pct_a_mult, logfc.threshold = input$logfc_a_mult, assay = "RNA", test.use = input$test.use_a_mult)
ips.markers_mult_b <- FindMarkers(v$scDatat_subset, ident.1 = input$gene1_mult, ident.2 = input$gene2_mult, only.pos = F, min.pct = input$min_pct_a_mult, logfc.threshold = input$logfc_a_mult, assay = "RNA", test.use = input$test.use_a_mult)
v$ips.markers_mult_a <- ips.markers_mult_a
v$ips.markers_mult_b <- ips.markers_mult_b
shinyalert("Pairwise DEGs done", "Pairwise DEGs done, please run GSEA Analysis", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$volcano_mult.plot <- renderPlot({
if(is.null(v$ips.markers_mult_b)){
return(NULL)
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
EnhancedVolcano(toptable = v$ips.markers_mult_b, lab = row.names(v$ips.markers_mult_b), x ="avg_log2FC", y ="p_val_adj", pointSize = 1, labSize = 5, legendLabSize = 12, axisLabSize = 12)
})
}
})
output$dega_mult.plot <- renderPlot({
if(is.null(v$ips.markers_mult_a)){
return(NULL)
}else{
withProgress(message="Generating Heatmap...", value=0, {
v$ips.markers_mult_a %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scDatat_subset, features = top10$gene, assay = "RNA", size = 5, angle = 45) + theme(axis.text.y = element_text(size = 8)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
output$gsea.ct_mult1.select <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
selectInput("gsea_mult.ct1", label = "Celltype1",
choices = as.vector(v$scDatat$primary.predict))
}
})
output$gsea.ct_mult2.select <- renderUI({
if(is.null(v$scDatat)){
return(NULL)
}else{
selectInput("gsea_mult.ct2", label = "Celltype2",
choices = as.vector(v$scDatat$primary.predict))
}
})
observeEvent(input$gsea_mult, {
tpmFiles2 <- v$scDatat
if (is.null(tpmFiles2)){
v$scDatat <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating gene set enrichment analysis...", value=0, {
if(input$species_gsea_mult == "Homo sapiens" & input$category_gsea_mult == "H"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_hs_go_mult <- msigdbr(species = "Homo sapiens", category = "H")
print(v$msigdbr_hs_go_mult)
v$pathways_mult <- split(x = v$msigdbr_hs_go_mult$gene_symbol, f = v$msigdbr_hs_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Mus musculus" & input$category_gsea_mult == "H"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_mm_go_mult <- msigdbr(species = "Mus musculus", category = "H")
print(v$msigdbr_mm_go_mult)
v$pathways_mult <- split(x = v$msigdbr_mm_go_mult$gene_symbol, f = v$msigdbr_mm_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Homo sapiens" & input$category_gsea_mult == "C2"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_hs_go_mult <- msigdbr(species = "Homo sapiens", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_hs_go_mult)
v$pathways_mult <- split(x = v$msigdbr_hs_go_mult$gene_symbol, f = v$msigdbr_hs_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Mus musculus" & input$category_gsea_mult == "C2"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_mm_go_mult <- msigdbr(species = "Mus musculus", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_mm_go_mult)
v$pathways_mult <- split(x = v$msigdbr_mm_go_mult$gene_symbol, f = v$msigdbr_mm_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Homo sapiens" & input$category_gsea_mult == "C5"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_hs_go_mult <- msigdbr(species = "Homo sapiens", category = "C5")
print(v$msigdbr_hs_go_mult)
v$pathways_mult <- split(x = v$msigdbr_hs_go_mult$gene_symbol, f = v$msigdbr_hs_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Mus musculus" & input$category_gsea_mult == "C5"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_mm_go_mult <- msigdbr(species = "Mus musculus", category = "C5")
print(v$msigdbr_mm_go_mult)
v$pathways_mult <- split(x = v$msigdbr_mm_go_mult$gene_symbol, f = v$msigdbr_mm_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Homo sapiens" & input$category_gsea_mult == "C7"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_hs_go_mult <- msigdbr(species = "Homo sapiens", category = "C7")
print(v$msigdbr_hs_go_mult)
v$pathways_mult <- split(x = v$msigdbr_hs_go_mult$gene_symbol, f = v$msigdbr_hs_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Mus musculus" & input$category_gsea_mult == "C7"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_mm_go_mult <- msigdbr(species = "Mus musculus", category = "C7")
print(v$msigdbr_mm_go_mult)
v$pathways_mult <- split(x = v$msigdbr_mm_go_mult$gene_symbol, f = v$msigdbr_mm_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Homo sapiens" & input$category_gsea_mult == "C8"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_hs_go_mult <- msigdbr(species = "Homo sapiens", category = "C8")
print(v$msigdbr_hs_go_mult)
v$pathways_mult <- split(x = v$msigdbr_hs_go_mult$gene_symbol, f = v$msigdbr_hs_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
if(input$species_gsea_mult == "Mus musculus" & input$category_gsea_mult == "C8"){
DefaultAssay(v$scDatat) <- "RNA"
v$msigdbr_mm_go_mult <- msigdbr(species = "Mus musculus", category = "C8")
print(v$msigdbr_mm_go_mult)
v$pathways_mult <- split(x = v$msigdbr_mm_go_mult$gene_symbol, f = v$msigdbr_mm_go_mult$gs_name)
print(v$pathways_mult)
v$markers_mult <- FindMarkers(v$scDatat, ident.1 = input$gsea_mult.ct1, ident.2 = input$gsea_mult.ct2, min.pct = input$min_pct_mult1, logfc.threshold = input$logfc_mult1, test.use = input$test.use_mult1)
v$markers_mult <- v$markers_mult %>% arrange(desc(avg_log2FC))
print(v$markers_mult)
v$markers_mult.log2FC <- v$markers_mult$avg_log2FC
names(v$markers_mult.log2FC) <- row.names(v$markers_mult)
v$markers_mult.log2FC <- sort(na.omit(v$markers_mult.log2FC), decreasing = TRUE)
print(v$markers_mult.log2FC)
v$fgseaRes_mult <- fgsea(pathways = v$pathways_mult, stats = v$markers_mult.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_mult)
v$topPathwaysUp_mult <- v$fgseaRes_mult[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_mult <- v$fgseaRes_mult[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_mult <- c(v$topPathwaysUp_mult, rev(v$topPathwaysDown_mult))
}
output$gsea_mult.done <- renderText(paste0("Gene set enrichment done!"))
v$isGSEAmultdone <- TRUE
})
}
})
output$gsea_mult.select <- renderUI({
if(is.null(v$pathways_mult)){
return(NULL)
}else{
selectInput("gsea_mult.pathway", label = "Gene set to visualise",
choices = names(v$pathways_mult))
}
})
output$gsea_mult_plot <- renderPlot({
if(is.null(v$pathways_mult) || is.null(v$isGSEAmultdone)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotEnrichment(v$pathways_mult[[input$gsea_mult.pathway]], v$markers_mult.log2FC) + labs(title=input$gsea_mult.pathway)
})
}
})
output$gsea_mult_plot1 <- renderPlot({
if(is.null(v$pathways_mult) || is.null(v$isGSEAmultdone)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotGseaTable(v$pathways_mult[v$topPathways_mult], v$markers_mult.log2FC, v$fgseaRes_mult, gseaParam=0.5)
})
}
})
output$gsea_mult.table <- DT::renderDataTable(
v$fgseaRes_mult, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_gsea_mult.table <- downloadHandler(
filename = function(){"GSEA Results.csv"},
content = function(fname){
withProgress(message="Downloading GSEA Results...", value=0, {
fwrite(v$fgseaRes_mult, fname)
})
}
)
observe({if(input$scAnalysis_type == "Multiome"){
shinyFileChoose(
input,
'dir_multi_atac',
roots = c(home = '.'),
filetypes = c('', 'txt', 'bigWig', "tsv", "csv", "bw", "gz", "tbi")
)
dir_multi_atac <- reactive(input$dir_multi_atac)
output$dir_multi_atac <- renderPrint({ # use renderText instead of renderPrint
as.character(parseFilePaths(c(home = '.'), dir_multi_atac())$datapath)
})
}
})
output$countdataDT2c <- renderDataTable({
if(!is.null(v$scDatan.rna))
{
if(ncol(v$scDatan.rna) > 20 )
return(as.matrix(v$scDatan@assays$RNA@counts[,1:20]))
}
}, server = FALSE)
output$countdataDT2d <- renderDataTable({
if(!is.null(v$scDatan.atac))
{
if(ncol(v$scDatan.atac) > 20 )
return(as.matrix(v$scDatan@assays$ATAC@counts[,1:20]))
}
}, server = FALSE)
observeEvent(input$filter_seurat2b, {
tpmFiles3 <- input$tpmFiles3
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$scDatan)
v$scDatan <- subset(v$scDatan, subset = nFeature_RNA > input$obsa & nFeature_RNA < input$obsa1 & nCount_ATAC > input$obsa3 & nCount_ATAC < input$obsa4 & percent.mt < input$obsa2)
print(v$scDatan)
shinyalert("Data processed", "Data processed, please view the Violin Plots", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
observeEvent(input$doSCTransform_multi, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running scTransform...", value = 0,{
incProgress(0.5, message = "Running scTransform...")
DefaultAssay(v$scDatan) <- "RNA"
v$scDatan <- SCTransform(v$scDatan, variable.features.n = input$var.genes_mult, vars.to.regress = "percent.mt", verbose = FALSE, conserve.memory = T)
incProgress(0.5)
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scDatan)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scDatan), 10)
variable_feature1 <- VariableFeaturePlot(v$scDatan)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("scTransform done", "scTransform done, please perform PCA", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
output$VarGenes_mult1 <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scDatan)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Multiomics_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scDatan)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scDatan@assays$RNA@var.features, file = txtfile)
})
}
})
})
}
})
observeEvent(input$runPCA_mult_seurat1, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0,{
incProgress(0.5, message = "Running PCA...")
v$scDatan <- RunPCA(v$scDatan, verbose = FALSE)
print(v$scDatan[["pca"]], dims = 1:5, nfeatures = 5)
v$isPCAdone1 <- TRUE
PCA_plot1a <- DimPlot(v$scDatan, reduction = "pca", label = T)
print(PCA_plot1a)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scDatan)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
shinyalert("PCA performed", "PCA performed, please perform clustering", type = "success", imageWidth = 10, imageHeight = 10)
}
)}
})
output$PCAplot_mult_seurat_h5_2 <- renderPlotly({
if(is.null(v$isPCAdone1)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of multiomics dataset...", value=0, {
DimPlot(v$scDatan, reduction = "pca", label = T)
})
}
})
output$vizPlot_mult_seurat1 <- renderPlot({
if(is.null(v$scDatan)){
return(NULL)
}else{
VizDimLoadings(v$scDatan, dims = as.numeric(input$select.pc_mult_seurat))
}
})
output$PCHeatmap_mult_seurat1 <- renderPlot({
if(is.null(v$scDatan)){
return(NULL)
}else{
DimHeatmap(v$scDatan, dims = as.numeric(input$select.pc_mult_seurat1))
}
})
output$PCtable_mult_seurat1 <- DT::renderDataTable({
if(is.null(v$scDatan) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_mult_seurat1 <- renderPlot({
if(is.null(v$scDatan)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scDatan)
})
}
}, height = 400, width = 450)
observeEvent(input$findCluster_mult_seurat1, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
v$scDatan <- FindNeighbors(v$scDatan, reduction = "pca", dims = 1:input$dim.used_mult_seurat1)
v$scDatan <- FindClusters(v$scDatan, resolution = input$clus.res_mult_seurat1)
#output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterdone <- TRUE
shinyalert("Clustering performed", "Clustering performed, please run UMAP", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Cluster2DPlot_mult_seurat1 <- renderPlotly({
if(is.null(v$isClusterdone)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scDatan, reduction = "pca", label = T)
})
}
})
observeEvent(input$runUMAP_mult_seurat1, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0,{
incProgress(0.5, message = "Running UMAP...")
v$scDatan <- RunUMAP(v$scDatan, reduction.name = "umap.rna", dims = 1:input$dim.used_mult_seurat1, reduction.key = 'rnaUMAP_')
print(v$scDatan)
DefaultAssay(v$scDatan) <- "ATAC"
v$scDatan <- RunTFIDF(v$scDatan)
v$scDatan <- FindTopFeatures(v$scDatan)
v$scDatan <- RunSVD(v$scDatan)
v$scDatan <- RunUMAP(v$scDatan, reduction = 'lsi', dims = 2:input$dim.used_mult_seurat1, reduction.name = "umap.atac", reduction.key = "atacUMAP_")
v$scDatan <- FindMultiModalNeighbors(v$scDatan, reduction.list = list("pca", "lsi"), dims.list = list(1:input$dim.used_mult_seurat1, 2:input$dim.used_mult_seurat1))
v$scDatan <- FindClusters(v$scDatan, resolution = input$clus.res_mult_seurat2b, graph.name = "wsnn", algorithm = 3, verbose = FALSE)
v$scDatan <- RunUMAP(v$scDatan, dims = 1:input$dim.used_mult_seurat1, nn.name = "weighted.nn", reduction.name = "wnn.umap", reduction.key = "wnnUMAP_")
v$isUMAPdone <- TRUE
print(v$scDatan)
UMAP_plot1a <- DimPlot(v$scDatan, reduction = "umap.rna", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("RNA")
UMAP_plot1b <- DimPlot(v$scDatan, reduction = "umap.atac", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("ATAC")
UMAP_plot1c <- DimPlot(v$scDatan, reduction = "wnn.umap", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("WNN")
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
shinyalert("UMAP done", "UMAP done, please run tSNE", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$UMAPplot1_a <- renderPlotly({
if(is.null(v$isUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatan, reduction = "umap.rna", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("RNA")
})
}
})
output$UMAPplot1_b <- renderPlotly({
if(is.null(v$isUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatan, reduction = "umap.atac", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("ATAC")
})
}
})
output$UMAPplot1_c <- renderPlotly({
if(is.null(v$isUMAPdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
DimPlot(v$scDatan, reduction = "wnn.umap", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("WNN")
})
}
})
observeEvent(input$runTSNE_mult_seurat1, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running TSNE...", value = 0,{
incProgress(0.5, message = "Running TSNE...")
v$scDatan <- RunTSNE(v$scDatan, dims = 1:input$dim.used_mult_seurat1, reduction.name = 'tsne.rna', reduction.key = 'rnaTSNE_')
DefaultAssay(v$scDatan) <- "ATAC"
v$scDatan <- RunTSNE(v$scDatan, reduction = 'lsi', dims = 2:input$dim.used_mult_seurat1, reduction.name = "tsne.atac", reduction.key = "atacTSNE_")
v$scDatan <- RunTSNE(v$scDatan, dims = 1:input$dim.used_mult_seurat1, nn.name = "weighted.nn", reduction.name = "wnn.tsne", reduction.key = "wnnTSNE_")
v$isTSNEdone <- TRUE
UMAP_plot1a <- DimPlot(v$scDatan, reduction = "tsne.rna", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("RNA")
UMAP_plot1b <- DimPlot(v$scDatan, reduction = "tsne.atac", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("ATAC")
UMAP_plot1c <- DimPlot(v$scDatan, reduction = "wnn.tsne", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("WNN")
print(UMAP_plot1a)
print(UMAP_plot1b)
print(UMAP_plot1c)
shinyalert("tSNE done", "tSNE done, please run Cell type identification", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$TSNEplot1_a <- renderPlotly({
if(is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating TSNE Plot of integrated dataset...", value=0, {
DimPlot(v$scDatan, reduction = "tsne.rna", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("RNA")
})
}
})
output$TSNEplot1_b <- renderPlotly({
if(is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating TSNE Plot of integrated dataset...", value=0, {
DimPlot(v$scDatan, reduction = "tsne.atac", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("ATAC")
})
}
})
output$TSNEplot1_c <- renderPlotly({
if(is.null(v$isTSNEdone)){
plotly_empty()
}else{
withProgress(message="Generating TSNE Plot of integrated dataset...", value=0, {
DimPlot(v$scDatan, reduction = "wnn.tsne", label = TRUE, label.size = 2.5, repel = TRUE) + ggtitle("WNN")
})
}
})
observeEvent(input$doCELLiD_multiome_seurat, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running CELLiD...", value = 0.3, {
ref = readRDS(url("https://www.immunesinglecell.org/api/vishuo/download/getCellidRef"))
v$scDatan.rna.data.average = AverageExpression(v$scDatan)
v$scDatan.rna.data.average = round(v$scDatan.rna.data.average$RNA, 2)
if(input$cellatlas_mult_multiome_seurat == "all"){
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, ref)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, adipose1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, adrenal_gland1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, blood1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, bone_marrow1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, brain1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, breast1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, breast_milk1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, eye1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, gut1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, heart1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, kidney1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, liver1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, lung1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, pancreas1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, PDAC1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, skin1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, testis1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, thymus1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_mult_multiome_seurat == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
v$res = FastIntegration::CELLiD(v$scDatan.rna.data.average, tonsil1)
print(v$res)
v$scDatan$primary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),1]
v$scDatan$secondary.predict = v$res[as.numeric(v$scDatan$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(v$res) <- newheaders
print(v$scDatan@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(v$res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
v$isCELLiDMultiomedone <- TRUE
shinyalert("Cell type identification done", "Cell type identification done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$Umap_cellid_multiome_seurat <- renderPlotly({
if(is.null(v$scDatan) || is.null(v$isCELLiDMultiomedone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scDatan, reduction = input$assay_mult_multiome_seurat, group.by = "primary.predict", label = T, label.size = 3)
})
}
})
output$Umap_cellid_multiome_seurat1 <- renderPlotly({
if(is.null(v$scDatan) || is.null(v$isCELLiDMultiomedone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP from CELLiD...", value=0, {
DimPlot(v$scDatan, reduction = input$assay_mult_multiome_seurat, group.by = "secondary.predict", label = T, label.size = 3)
})
}
})
output$ct_multiome_seurat.table <- DT::renderDataTable(
v$res, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$download_cellid_multiome_seurat_prediction <- downloadHandler(
filename = function(){"CELLiD predictions.csv"},
content = function(fname){
withProgress(message="Downloading CELLiD predictions...", value=0, {
write.csv(v$res, fname)
})
}
)
observeEvent(input$Vis3_multiome, {
tpmFiles3 <- input$tpmFiles3
tpmFiles3 <- v$scDatan
if (is.null(tpmFiles3)){
v$scDatan <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Visualizing...", value = 0,{
incProgress(0.5, message = "Visualizing...")
DefaultAssay(v$scDatan) <- "RNA"
})
}
})
output$vis.gene.select_multiome <- renderUI({
if(is.null(v$scDatan)){
return(NULL)
}else{
selectInput("vis.gene_multiome", label = "Gene to visualise",
choices = rownames(v$scDatan[[input$assay_multiome_seurat2a]]))
}
})
output$vis.plot_multiome <- renderPlotly({
if(is.null(v$scDatan)){
return(NULL)
}else{
withProgress(message="Generating Feature Plot...", value=0, {
FeaturePlot(v$scDatan, input$vis.gene_multiome, cols = c("lightgrey", "darkgreen"), order = T, reduction = input$assay_multiome_seurat1a)
})
}
})
output$vis.gene.select_multiome1 <- renderUI({
if(is.null(v$scDatan)){
return(NULL)
}else{
selectInput("vis.gene_multiome1", label = "Gene to visualise",
choices = rownames(v$scDatan[["ATAC"]]))
}
})
output$vis1.plot_multiome <- renderPlotly({
if(is.null(v$scDatan)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(v$scDatan, input$vis.gene_multiome1, cols = c("lightgrey", "darkgreen"), order = T, reduction = "umap.atac")
})
}
})
##---------------Spatial Transcriptomics Analysis-------------------##
output$spatial_image <- renderImage({
list(src = "www/spatial_fig.png",
height = 275, width=1000)
}, deleteFile = FALSE)
#volumes <- getVolumes()
shinyDirChoose(
input,
'dir',
roots=c(home = '.'),
filetypes = c('', 'txt', 'bigWig', "tsv", "csv", "bw")
)
dir <- reactive(input$dir)
output$dir <- renderPrint({ # use renderText instead of renderPrint
parseDirPath(c(home = '.'), dir())
})
observeEvent(input$loadexample_seurat_spatial, {
withProgress(message="Loading example Data...", value=0.5, {
path <- "spatial/V1_Breast_Cancer_Block_A_Section_1/"
sp.data <- Load10X_Spatial(data.dir = path, filename = "filtered_feature_bc_matrix.h5")
sp.data[["percent.mt"]] <- PercentageFeatureSet(sp.data, pattern = "^MT-")
})
if (is.null(sp.data)){
v$scData_spatial <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(sp.data)
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_seurat_spatial", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData_spatial <- sp.data
})
}
})
observeEvent(input$loadButton3, {
if(input$scInput3 == "SpaceRanger output" & input$scAnalysis_sp == "Seurat"){
tpmFile_spatial <- input$tpmFile_spatial
tpmFiles_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFile_spatial$datapath)
print(tpmFile_spatial$name)
print(file.exists(paste(tpmFile_spatial$datapath[1], "/", tpmFile_spatial$name[1], sep="")))
exp.data_spatial <- Load10X_Spatial(parseDirPath(c(home = '.'), dir()), filename = tpmFile_spatial$name, assay = "Spatial")
additional.ident <- NULL
incProgress(0.5, "Creating Seurat Object")
v$scData_spatial <- exp.data_spatial
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
}
)
observeEvent(input$reset3, {
session$reload()
print("Reset done")
})
observeEvent(input$loadButton3a, {
if(input$scInput3 == "SpaceRanger output" & input$scAnalysis_sp == "GraphST"){
sc <- reticulate::import("scanpy", convert = FALSE)
#scvi <- import("scvi", convert = FALSE)
scipy <- reticulate::import("scipy", convert = FALSE)
pot <- reticulate::import("ot", convert = FALSE)
pd <- reticulate::import("pandas", convert = F)
gt <- reticulate::import("GraphST", convert = F)
matplotlib <- reticulate::import ("matplotlib", convert = F)
gt1 <- reticulate::import("GraphST.preprocess", convert = F)
gt2 <- reticulate::import("GraphST.utils", convert = F)
#torch <- reticulate::import("torch", convert = FALSE)
np <- reticulate::import("numpy", convert = F)
sns <- reticulate::import("seaborn", convert = F)
skmisc <- reticulate::import("skmisc", convert = F)
#device = torch$device('cuda:0')
tpmFile_graphst <- input$tpmFile_graphst
if (is.null(tpmFile_graphst)){
v$scData_graphst <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFile_graphst$datapath)
print(tpmFile_graphst$name)
print(file.exists(paste(tpmFile_graphst$datapath[1], "/", tpmFile_graphst$name[1], sep="")))
file_fold <- parseDirPath(c(home = '.'), dir_graphst())
adata <- sc$read_visium(file_fold, count_file=tpmFile_graphst$name, load_images=TRUE)
print(adata)
adata$var_names_make_unique()
adata$obs_names_make_unique()
additional.ident <- NULL
incProgress(0.5, "Creating Seurat Object")
v$scData_spatial <- exp.data_spatial
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
}
)
observeEvent(input$reset3a, {
session$reload()
print("Reset done")
})
output$countdataDT_spatial <- renderDataTable({
if(!is.null(v$scData_spatial))
{
if(ncol(v$scData_spatial) > 20 )
return(as.matrix(v$scData_spatial@assays$Spatial@counts[,1:20]))
}
}, server = FALSE)
output$h_e <- renderPlot({
if(is.null(v$scData_spatial)){
plotly_empty()
}else{
print(SpatialDimPlot(v$scData_spatial, pt.size.factor = 0) + NoLegend())
}
})
observeEvent(input$loadexample_graphst_spatial, {
withProgress(message = "Loading example data...", value = 0.3, {
if(input$scAnalysis_sp == "GraphST"){
sc <- reticulate::import("scanpy", convert = FALSE)
#scvi <- import("scvi", convert = FALSE)
scipy <- reticulate::import("scipy", convert = FALSE)
pot <- reticulate::import("ot", convert = FALSE)
pd <- reticulate::import("pandas", convert = F)
gt <- reticulate::import("GraphST", convert = F)
matplotlib <- reticulate::import ("matplotlib", convert = F)
gt1 <- reticulate::import("GraphST.preprocess", convert = F)
gt2 <- reticulate::import("GraphST.utils", convert = F)
np <- reticulate::import("numpy", convert = F)
sns <- reticulate::import("seaborn", convert = F)
#torch <- reticulate::import("torch", convert = F)
skmisc <- reticulate::import("skmisc", convert = F)
#device = torch$device('cuda:0')
withProgress(message="Loading and Processing Data...", value=0, {
adata <- sc$read_visium('spatial/V1_Breast_Cancer_Block_A_Section_1/', count_file='filtered_feature_bc_matrix.h5', load_images=TRUE)
print(adata)
adata$var_names_make_unique()
adata$obs_names_make_unique()
#sce = zellkonverter::AnnData2SCE(adata = adata, layers = T)
#counts(sce) <- assay(sce, "X")
#sce <- logNormCounts(sce)
#spatial <- as.Seurat(x = sce, counts = "X", data = "logcounts")
sp.data <- Load10X_Spatial(data.dir = 'spatial/V1_Breast_Cancer_Block_A_Section_1/', filename = "filtered_feature_bc_matrix.h5")
sp.data[["percent.mt"]] <- PercentageFeatureSet(sp.data, pattern = "^MT-")
v$scData_spatial <- sp.data
v$scData_spatial_graphst <- adata
#print(v$scData_spatial1)
#print(v$scData_spatial_graphst)
})
}
})
})
output$h_e_graphst <- renderPlot({
if(is.null(v$scData_spatial)){
plotly_empty()
}else{
print(SpatialDimPlot(v$scData_spatial, pt.size.factor = 0) + NoLegend())
}
})
output$countdataDT_spatial_graphst <- renderDataTable({
if(!is.null(v$scData_spatial))
{
if(ncol(v$scData_spatial) > 20 )
return(as.matrix(v$scData_spatial@assays$Spatial@counts[,1:20]))
}
}, server = FALSE)
output$countdataDT_xenium <- renderDataTable({
if(!is.null(v$scData_xenium))
{
if(ncol(v$scData_xenium) > 20 )
return(as.matrix(v$scData_xenium@assays$Xenium@counts[,1:20]))
}
}, server = FALSE)
output$h_e_xenium <- renderPlot({
if(is.null(v$scData_xenium)){
plotly_empty()
}else{
print(ImageDimPlot(v$scData_xenium, fov = "fov") + NoLegend())
}
})
observeEvent(input$filter_spatial, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Filter low quality cells...", value = 0.3, {
v$scData_spatial <- subset(v$scData_spatial, subset = nCount_Spatial > input$obs & nFeature_Spatial > input$obs1 & percent.mt < input$obs2)
})
}
#}
})
output$nCount_SpatialPlot <- renderPlot({
if(is.null(v$scData_spatial)){
plotly_empty()
}else{
VlnPlot(v$scData_spatial, features = c("nCount_Spatial","nFeature_Spatial","percent.mt"), pt.size = 0.1) + NoLegend()
}
})
output$SpatialFeaturePlot <- renderPlot({
if(is.null(v$scData_spatial)){
plotly_empty()
}else{
SpatialFeaturePlot(v$scData_spatial, features = c("nCount_Spatial","nFeature_Spatial","percent.mt")) + theme(legend.position = "top")
}
})
observeEvent(input$doSCTransform_sp, {
if(input$scInput3 == "SpaceRanger output" & input$scAnalysis_sp == "Seurat" | input$scAnalysis_sp == "GraphST"){
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running scTransform...", value = 0,{
incProgress(0.5, message = "Running scTransform...")
v$scData_spatial <- SCTransform(v$scData_spatial, variable.features.n = input$var.genes_sp, assay = "Spatial", verbose = FALSE, conserve.memory = T)
incProgress(0.5)
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scData_spatial)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scData_spatial), 10)
variable_feature1 <- VariableFeaturePlot(v$scData_spatial)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("scTransform done", "scTransform done, please visualize gene expression", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
output$VarGenes_sp <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scData_spatial)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Spatial_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scData_spatial)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scData_spatial@assays$Spatial@var.features, file = txtfile)
})
}
})
})
}
}
})
#observeEvent(input$doSCTransform_sp, {
# if(input$scInput3 == "SpaceRanger output" & input$scAnalysis_sp == "Seurat" | input$scAnalysis_sp == "GraphST"){
# withProgress(message = "Running scTransform...", value = 0,{
# incProgress(0.5, message = "Running scTransform...")
# v$scData_spatial <- SCTransform(v$scData_spatial, variable.features.n = input$var.genes_sp, assay = "Spatial", verbose = FALSE, conserve.memory = T)
# incProgress(0.5)
# #VarGeneText <- paste0("Number of variable genes: ", length(v$scData@assays$RNA@var.features))
# #output$nVarGenes <- renderText(VarGeneText)
# varGenePlotInput <- function(){
# if(is.null(v$scData_spatial)){
# return(NULL)
# }else{
# withProgress(message="Plotting variable genes...", value=0, {
# top10 <- head(VariableFeatures(v$scData_spatial), 10)
# variable_feature1 <- VariableFeaturePlot(v$scData_spatial)
# variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
# print (variable_feature1)
# print (variable_feature2)
# shinyalert("scTransform done", "scTransform done, please visualize gene expression", type = "success", imageWidth = 10, imageHeight = 10)
# })
# }
# }
# output$VarGenes_sp <- renderPlot({
# varGenePlotInput()
# }, height = 800, width = 850)
# observeEvent(input$PDFc, {
# if(!is.null(v$scData_spatial)){
# withProgress(message="Downloading plot PDF files...", value=0, {
# print(getwd())
# pdfDir <- paste0(getwd(), .Platform$file.sep, "Spatial_results/Generated_reports_", Sys.Date())
# if(!dir.exists(pdfDir)){
# dir.create(pdfDir)
# }
# filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
# i = 0
# while(file.exists(filename2)){
# filename2 <- paste0(pdfDir, .Platform$file.sep,
# "Var_genes_plot_",
# Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
# i = i + 1;
# }
# prePlot()
# pdf(filename2,
# width=as.numeric(input$pdf_w),
# height=as.numeric(input$pdf_h))
# plot1 <- VariableFeaturePlot(v$scData_spatial)
# print(plot1)
# dev.off()
# txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
# txtfile <- sub(".pdf", ".txt", txtfile)
# write(v$scData_spatial@assays$Spatial@var.features, file = txtfile)
# })
# }
# })
# })
# }
#})
observeEvent(input$doSCTransform_xenium, {
if(input$scAnalysis_platform == "Xenium"){
tpmFile_xenium <- v$scData_xenium
if (is.null(tpmFile_xenium)){
v$scData_xenium <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running scTransform...", value = 0,{
incProgress(0.5, message = "Running scTransform...")
v$scData_xenium <- SCTransform(v$scData_xenium, variable.features.n = input$var.genes_xenium, assay = "Xenium", verbose = FALSE, conserve.memory = T)
incProgress(0.5)
#VarGeneText <- paste0("Number of variable genes: ", length(v$scData@assays$RNA@var.features))
#output$nVarGenes <- renderText(VarGeneText)
varGenePlotInput <- function(){
if(is.null(v$scData_xenium)){
return(NULL)
}else{
withProgress(message="Plotting variable genes...", value=0, {
top10 <- head(VariableFeatures(v$scData_xenium), 10)
variable_feature1 <- VariableFeaturePlot(v$scData_xenium)
variable_feature2 <- LabelPoints(plot = variable_feature1, points = top10, repel = TRUE)
print (variable_feature1)
print (variable_feature2)
shinyalert("scTransform done", "scTransform done, please visualize gene expression", type = "success", imageWidth = 10, imageHeight = 10)
})
}
}
output$VarGenes_xenium <- renderPlot({
varGenePlotInput()
}, height = 500, width = 600)
observeEvent(input$PDFc, {
if(!is.null(v$scData_xenium)){
withProgress(message="Downloading plot PDF files...", value=0, {
print(getwd())
pdfDir <- paste0(getwd(), .Platform$file.sep, "Spatial_results/Generated_reports_", Sys.Date())
if(!dir.exists(pdfDir)){
dir.create(pdfDir)
}
filename2 <- paste0(pdfDir, .Platform$file.sep,"Var_genes_plot_", Sys.Date(), ".pdf")
i = 0
while(file.exists(filename2)){
filename2 <- paste0(pdfDir, .Platform$file.sep,
"Var_genes_plot_",
Sys.Date(), "_", sprintf("%03d", i + 1), ".pdf");
i = i + 1;
}
prePlot()
pdf(filename2,
width=as.numeric(input$pdf_w),
height=as.numeric(input$pdf_h))
plot1 <- VariableFeaturePlot(v$scData_xenium)
print(plot1)
dev.off()
txtfile <- sub("Var_genes_plot_", "Var_gene_list_", filename2)
txtfile <- sub(".pdf", ".txt", txtfile)
write(v$scData_xenium@assays$Xenium@var.features, file = txtfile)
})
}
})
})
}
}
})
output$sp.gene.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
selectInput("sp.gene", label = "Visualize normalized expression",
choices = rownames(v$scData_spatial[["Spatial"]]))
}
})
output$sp.plot <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
withProgress(message="Generating Feature Plot...", value=0, {
SpatialFeaturePlot(v$scData_spatial, features = input$sp.gene)
})
}
})
observeEvent(input$runPCA_spatial, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0,{
incProgress(0.5, message = "Running PCA...")
v$scData_spatial <- RunPCA(v$scData_spatial, assay = "SCT", verbose = FALSE)
#DefaultAssay(v$scDatan) <- "SCT"
print(v$scData_spatial[["pca"]], dims = 1:5, nfeatures = 5)
v$isPCAdone <- TRUE
PCA_plot1a <- DimPlot(v$scData_spatial, reduction = "pca", label = T)
print(PCA_plot1a)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scData_spatial)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
}
)}
})
output$PCAplot_spatial <- renderPlotly({
if(is.null(v$isPCAdone)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of spatial dataset...", value=0, {
DimPlot(v$scData_spatial, reduction = "pca", label = T)
})
}
})
output$vizPlot_spatial <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
VizDimLoadings(v$scData_spatial, dims = as.numeric(input$select.pc_spatial))
}
})
output$PCHeatmap_spatial <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
DimHeatmap(v$scData_spatial, dims = as.numeric(input$select.pc_spatial), assays = "SCT")
}
})
output$PCtable_spatial <- DT::renderDataTable({
if(is.null(v$scData_spatial) ){
return(NULL)
}else{
v$pcGenes
}
}, server = FALSE, options = list(scrollX = TRUE))
output$Elbow_spatial <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scData_spatial, ndims = 50)
})
}
}, height = 400, width = 450)
observeEvent(input$runPCA_xenium, {
tpmFile_xenium <- v$scData_xenium
if (is.null(tpmFile_xenium)){
v$scData_xenium <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running PCA...", value = 0,{
incProgress(0.5, message = "Running PCA...")
v$scData_xenium <- RunPCA(v$scData_xenium, npcs = 30, features = rownames(v$scData_xenium))
#DefaultAssay(v$scDatan) <- "SCT"
print(v$scData_xenium[["pca"]], dims = 1:5, nfeatures = 5)
v$isPCAXeniumdone <- TRUE
PCA_plot1a <- DimPlot(v$scData_xenium, reduction = "pca", label = T)
print(PCA_plot1a)
incProgress(0.4, message = "Getting list of PC genes...")
pc.table <- list()
for(i in 1:20){
pcg <- TopFeatures(v$scData_xenium)
pc.table[[i]] <- pcg
}
pc.table <- as.data.frame(pc.table, col.names = paste0("PC", 1:20))
v$pcGenes <- pc.table
}
)}
})
output$PCAplot_xenium <- renderPlotly({
if(is.null(v$isPCAXeniumdone)){
plotly_empty()
}else{
withProgress(message="Generating PCA Plot of spatial dataset...", value=0, {
DimPlot(v$scData_xenium, reduction = "pca", label = T)
})
}
})
output$Elbow_xenium <- renderPlot({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
withProgress(message="Generating Elbow Plot...", value=0.5, {
ElbowPlot(v$scData_xenium, ndims = 30)
})
}
}, height = 400, width = 450)
shinyDirChoose(
input,
'dir_graphst',
roots = c(home = '.'),
filetypes = c('', 'txt', 'bigWig', "tsv", "csv", "bw")
)
dir_graphst <- reactive(input$dir_graphst)
output$dir_graphst <- renderPrint({ # use renderText instead of renderPrint
parseDirPath(c(home = '.'), dir_graphst())
})
observeEvent(input$findCluster_spatial, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
if(input$scAnalysis_sp1 == "Seurat"){
v$scData_spatial <- FindNeighbors(v$scData_spatial, reduction = "pca", dims = 1:input$dim.used_spatial)
v$scData_spatial <- FindClusters(v$scData_spatial, resolution = input$clus.res_spatial, graph.name = "SCT_nn")
#output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterSpatialdone <- TRUE
}
})
}
})
output$Cluster2DPlot_spatial <- renderPlotly({
if(is.null(v$isClusterSpatialdone)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scData_spatial, reduction = "umap", label = T)
})
}
})
output$silhouette_seurat <- renderText({
if(is.null(v$scData_spatial)|| is.null(v$isClusterSpatialdone)){
return(NULL)
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
print(v$scData_spatial@reductions$pca@cell.embeddings)
print(v$scData_spatial@meta.data)
v$test <- paste("Silhouette Score:", mean(silhouette(x = as.numeric(x = as.factor(x = v$scData_spatial$seurat_clusters)), dist = dist(x = v$scData_spatial@reductions$pca@cell.embeddings))[, 3]))
print(v$test)
})
}
})
observeEvent(input$findCluster_graphst, {
tpmFile_spatial <- input$tpmFile_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Finding clusters...", value = 0.3, {
if(input$scAnalysis_sp1 == "GraphST"){
sc <- reticulate::import("scanpy", convert = FALSE)
#scvi <- import("scvi", convert = FALSE)
scipy <- reticulate::import("scipy", convert = FALSE)
pot <- reticulate::import("ot", convert = FALSE)
pd <- reticulate::import("pandas", convert = F)
gt <- reticulate::import("GraphST", convert = F)
matplotlib <- reticulate::import ("matplotlib", convert = F)
gt1 <- reticulate::import("GraphST.preprocess", convert = F)
gt2 <- reticulate::import("GraphST.utils", convert = F)
np <- reticulate::import("numpy", convert = F)
sns <- reticulate::import("seaborn", convert = F)
#torch <- reticulate::import("torch", convert = F)
skmisc <- reticulate::import("skmisc", convert = F)
#device = torch$device('cuda:0')
withProgress(message="Clustering data...", value=0, {
model = gt$GraphST$GraphST(adata = v$scData_spatial_graphst)
v$scData_spatial_graphst = model$train()
print(v$scData_spatial_graphst)
cluster <- gt$clustering(adata = v$scData_spatial_graphst , n_clusters = input$cluster_graphst, method = input$cluster_method_graphst)
v$isClusterSpatialdone2 <- TRUE
print(v$scData_spatial_graphst)
print(cluster)
v$scData_spatial_graphst_obs = v$scData_spatial_graphst$obs
v$scData_spatial_graphst_obs$to_csv('adata_obs.csv')
#v$scData_spatial_graphst$write_csvs(dirname = 'test/')
print(v$scData_spatial_graphst_obs)
meta <- read.csv('adata_obs.csv', header = T, row.names = 1)
v$scData_spatial <- AddMetaData(v$scData_spatial, metadata = meta)
print(v$scData_spatial)
})
}
})
}
})
output$Cluster2DPlot_graphst <- renderPlot({
if(is.null(v$isClusterSpatialdone2)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
SpatialDimPlot(v$scData_spatial, group.by = "mclust", label = T)
})
}
})
output$DimPlot_GraphST <- renderPlotly({
if(is.null(v$scData_spatial)|| is.null(v$isClusterSpatialdone2)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
DimPlot(v$scData_spatial, group.by = "mclust", label = T)
})
}
})
output$SpatialDimPlot_GraphST <- renderPlot({
if(is.null(v$scData_spatial) || is.null(v$isClusterSpatialdone2)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
SpatialDimPlot(v$scData_spatial, group.by = "mclust", label = TRUE)
})
}
})
output$silhouette_graphst <- renderText({
if(is.null(v$scData_spatial)|| is.null(v$isClusterSpatialdone2)){
return(NULL)
}else{
withProgress(message="Generating UMAP Plot of integrated dataset...", value=0, {
print(v$scData_spatial@reductions$pca@cell.embeddings)
print(v$scData_spatial@meta.data)
v$test <- paste("Silhouette Score:", mean(silhouette(x = as.numeric(x = as.factor(x = v$scData_spatial$mclust)), dist = dist(x = v$scData_spatial@reductions$pca@cell.embeddings))[, 3]))
print(v$test)
})
}
})
observeEvent(input$findCluster_xenium, {
withProgress(message = "Finding clusters...", value = 0.3, {
v$scData_xenium <- FindNeighbors(v$scData_xenium, reduction = "pca", dims = 1:input$dim.used_xenium)
v$scData_xenium <- FindClusters(v$scData_xenium, resolution = input$clus.res_xenium)
#output$cluster1.done <- renderText(paste0("Clustering done!"))
v$isClusterXeniumdone <- TRUE
})
})
output$Cluster2DPlot_xenium <- renderPlotly({
if(is.null(v$isClusterXeniumdone)){
plotly_empty()
}else{
withProgress(message="Generating 2D Cluster Plot...", value=0, {
DimPlot(v$scData_xenium, reduction = "umap", label = T)
})
}
})
observeEvent(input$runUMAP_spatial, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0.3, {
v$scData_spatial <- RunUMAP(v$scData_spatial, dims = 1:input$dim.used_spatial, assay = "SCT", spread = 1)
v$isUMAPSpataildone <- TRUE
})
}
})
output$DimPlot_spatial <- renderPlotly({
if(is.null(v$scData_spatial) || is.null(v$isUMAPSpataildone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
DimPlot(v$scData_spatial, reduction = "umap", label = T)
})
}
})
output$SpatialDimPlot <- renderPlot({
if(is.null(v$scData_spatial) || is.null(v$isClusterSpatialdone)){
plotly_empty()
}else{
withProgress(message="Generating UMAP 2D Plot...", value=0, {
SpatialDimPlot(v$scData_spatial, label = TRUE, label.size = 3)
})
}
})
observeEvent(input$runUMAP_xenium, {
withProgress(message = "Running UMAP...", value = 0.3, {
v$scData_xenium <- RunUMAP(v$scData_xenium, dims = 1:input$dim.used_xenium, assay = "SCT", spread = 1)
v$isUMAPXeniumdone <- TRUE
})
})
output$umap_xenium <- renderPlotly({
if(is.null(v$scData_xenium) || is.null(v$isUMAPXeniumdone)){
plotly_empty()
}else{
DimPlot(v$scData_xenium)
}
})
output$xenium.gene3.select <- renderUI({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
selectInput("xenium.gene3", label = "Genes to visualize",
choices = rownames(v$scData_xenium@assays$Xenium@counts), selected = rownames(v$scData_xenium@assays$Xenium@counts)[20])
}
})
output$xenium_feature.plot <- renderPlotly({
if(is.null(v$scData_xenium) || is.null(v$isUMAPXeniumdone)){
plotly_empty()
}else{
FeaturePlot(v$scData_xenium, features = input$xenium.gene3)
}
})
output$spatialumap_xenium <- renderPlot({
if(is.null(v$scData_xenium) || is.null(v$isUMAPXeniumdone)){
plotly_empty()
}else{
ImageDimPlot(v$scData_xenium, cols = "polychrome", size = 0.75, molecules = input$xenium.gene3)
}
})
observeEvent(input$Vis_sp1, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Visualizing...", value = 0,{
incProgress(0.5, message = "Visualizing...")
DefaultAssay(v$scData_spatial) <- "Spatial"
#Idents(v$scData_spatial) <- "seurat_clusters"
})
}
})
output$sp.cluster.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
selectInput("sp.cluster", label = "Cluster to visualise",
choices = unique(v$scData_spatial@meta.data$seurat_clusters))
}
})
output$sp1.plot <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
withProgress(message="Generating Feature Plot...", value=0, {
SpatialDimPlot(v$scData_spatial, cells.highlight = CellsByIdentities(object = v$scData_spatial, idents = input$sp.cluster), facet.highlight = TRUE)
})
}
})
output$sp.cluster1.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
selectInput("sp.cluster1", label = "Cluster to visualise",
choices = unique(v$scData_spatial@meta.data$mclust))
}
})
output$sp2.plot <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
withProgress(message="Generating Feature Plot...", value=0, {
v$scData_spatial$mclust -> Idents(v$scData_spatial)
SpatialDimPlot(v$scData_spatial, cells.highlight = CellsByIdentities(object = v$scData_spatial, idents = input$sp.cluster1), facet.highlight = TRUE)
})
}
})
observeEvent(input$Vis_sp_xenium, {
withProgress(message = "Visualizing...", value = 0,{
incProgress(0.5, message = "Visualizing...")
DefaultAssay(v$scData_xenium) <- "Xenium"
#Idents(v$scData_spatial) <- "seurat_clusters"
})
})
output$xenium.cluster.select <- renderUI({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
selectInput("xenium.cluster", label = "Cluster to visualise",
choices = unique(v$scData_xenium@meta.data$seurat_clusters))
}
})
output$xenium1.plot <- renderPlot({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
withProgress(message="Generating Feature Plot...", value=0, {
ImageDimPlot(v$scData_xenium, size = 0.75, cells = colnames(subset(v$scData_xenium, seurat_clusters == input$xenium.cluster)))
})
}
})
observeEvent(input$loadexample_ref, {
withProgress(message="Loading example reference scRNA-seq data...", value=0.5, {
tpmFiles_ref <- read.csv('gene_counts_reference1.csv', header = T, row.names = 1, check.names = F)
meta_ref <- read.csv('metadata_reference1.csv', header = T, row.names = 1)
tpmFiles_ref <- CreateSeuratObject(tpmFiles_ref, meta.data = meta_ref)
tpmFiles_ref$cell_type = tpmFiles_ref$primary.predict
tpmFiles_ref$cell_type <- gsub(" ", "_", tpmFiles_ref$cell_type)
tpmFiles_ref$cell_type <- gsub("/", "_", tpmFiles_ref$cell_type)
tpmFiles_ref <- Seurat::SCTransform(tpmFiles_ref, ncells = 3000, verbose = FALSE)
tpmFiles_ref <- Seurat::RunPCA(tpmFiles_ref, verbose = FALSE)
tpmFiles_ref <- Seurat::RunUMAP(tpmFiles_ref, dims = 1:30, verbose = FALSE)
#tpmFiles_ref <- Seurat::FindNeighbors(tpmFiles_ref, dims = 1:30, verbose = FALSE)
#tpmFiles_ref <- Seurat::FindClusters(tpmFiles_ref, verbose = FALSE)
})
if (is.null(tpmFiles_ref)){
v$scRNAData <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_ref", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData1 <- tpmFiles_ref
})
}
})
observe({
if(input$loaduser_ref > 0){
print('1')
session$sendCustomMessage("myCallbackHandler2", "1")
}
})
output$scRNAPlot <- renderPlot({
if(is.null(v$scData1)){
plotly_empty()
}else{
withProgress(message="Generating scRNA-seq Plot...", value=0, {
DimPlot(v$scData1, group.by = "primary.predict", label = T, label.size = 3, repel = T) + NoLegend()
})
}
})
observeEvent(input$vis_spRNA, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Processing scRNA-seq dataset...", value = 0,{
incProgress(0.5, message = "Processing...")
spRNA_plot <- SpatialDimPlot(v$scData_spatial, label = TRUE, label.size = 3)
print (spRNA_plot)
#output$vis_sp.done <- renderText(paste0("Processing of scRNA-seq data done!"))
v$isVis_spRNAdone <- TRUE
})
}
})
output$spRNAPlot <- renderPlot({
if(is.null(v$isVis_spRNAdone)){
plotly_empty()
}else{
withProgress(message="Generating SpatialDim Plot...", value=0, {
SpatialDimPlot(v$scData_spatial, label = TRUE, label.size = 3)
})
}
})
observeEvent(input$doDeconv, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Performing deconvolution...", value = 0,{
incProgress(0.5, message = "Deconvoluting...")
v$anchors <- FindTransferAnchors(reference = v$scData1, query = v$scData_spatial, normalization.method = "SCT")
predictions.assay <- TransferData(anchorset = v$anchors, refdata = v$scData1$primary.predict, prediction.assay = TRUE,
weight.reduction = v$scData_spatial[["pca"]], dims = 1:30)
v$scData_spatial[["predictions"]] <- predictions.assay
print(v$scData_spatial@meta.data)
print(v$scData_spatial)
v$isDeconv_spRNAdone <- TRUE
#output$vis_sp.done <- renderText(paste0("Processing of scRNA-seq data done!"))
})
}
})
output$ct.select <- renderUI({
if(is.null(v$scData1)){
return(NULL)
}else{
selectInput("ct", label = "Celltype to visualise",
choices = unique(v$scData1$primary.predict))
}
})
output$DeconvPlot <- renderPlot({
if(is.null(v$scData_spatial) || is.null(v$isDeconv_spRNAdone)){
return(NULL)
}else{
withProgress(message="Generating SpatialDim Plot...", value=0, {
DefaultAssay(v$scData_spatial) <- "predictions"
SpatialFeaturePlot(v$scData_spatial, features = input$ct, pt.size.factor = 1.6, ncol = 2, crop = TRUE)
})
}
})
observeEvent(input$doDeconv_graphst, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Performing deconvolution...", value = 0,{
incProgress(0.5, message = "Deconvoluting...")
sc <- reticulate::import("scanpy", convert = FALSE)
#scvi <- import("scvi", convert = FALSE)
scipy <- reticulate::import("scipy", convert = FALSE)
pot <- reticulate::import("ot", convert = FALSE)
pd <- reticulate::import("pandas", convert = F)
gt <- reticulate::import("GraphST", convert = F)
matplotlib <- reticulate::import ("matplotlib", convert = F)
gt1 <- reticulate::import("GraphST.preprocess", convert = F)
gt2 <- reticulate::import("GraphST.utils", convert = F)
np <- reticulate::import("numpy", convert = F)
sns <- reticulate::import("seaborn", convert = F)
#torch <- reticulate::import("torch", convert = F)
skmisc <- reticulate::import("skmisc", convert = F)
#device = torch$device('cuda:0')
sceasy::convertFormat(v$scData1, from = "seurat", to = "anndata", outFile = 'reference.h5ad')
file_path = 'reference.h5ad' # Please replace 'file_path' with the scRNA download path.
#Reading reference data
v$adata_sc = sc$read(file_path)
v$adata_sc$var_names_make_unique()
#Pre-processing for ST data
gt$preprocess(v$scData_spatial_graphst)
gt$construct_interaction(v$scData_spatial_graphst)
gt$add_contrastive_label(v$scData_spatial_graphst)
#Pre-processing for reference data
gt$preprocess(v$adata_sc)
#Finding overlap genes between ST and reference data
v$test = gt1$filter_with_overlap_gene(adata = v$scData_spatial_graphst, adata_sc = v$adata_sc)
print(v$test[0])
print(v$test[1])
gt1$get_feature(v$test[0])
model = gt$GraphST$GraphST(v$test[0], v$test[1], deconvolution = T)
v$test = model$train_map()
print(v$test[0])
print("test")
print(v$test[1])
gt2$project_cell_to_spot(adata = v$test[0], adata_sc = v$test[1], retain_percent = 0.15)
v$test_obs = v$test[0]$obs
v$test_obs$to_csv('bdata_obs.csv')
#v$test[0]$write_csvs('ref1/')
meta1 <- read.csv('bdata_obs.csv', header = T, row.names = 1)
v$scData_spatial <- AddMetaData(v$scData_spatial, metadata = meta1)
})
}
})
output$graphst_ct.select <- renderUI({
if(is.null(v$scData1)){
return(NULL)
}else{
selectInput("ct_graphst", label = "Celltype to visualise",
choices = unique(v$scData1$cell_type))
}
})
output$DeconvPlot_graphst <- renderPlot({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
withProgress(message="Generating SpatialDim Plot...", value=0, {
#DefaultAssay(v$scData_spatial) <- "predictions"
SpatialFeaturePlot(v$scData_spatial, features = input$ct_graphst, pt.size.factor = 1.6, ncol = 2, crop = TRUE)
})
}
})
shinyDirChoose(
input,
'dir_xenium',
roots = c(home = '.'),
filetypes = c('', 'txt', 'bigWig', "tsv", "csv", "bw")
)
dir_xenium <- reactive(input$dir_xenium)
output$dir_xenium <- renderPrint({ # use renderText instead of renderPrint
parseDirPath(c(home = '.'), dir_xenium())
})
observeEvent(input$loadexample_xenium, {
withProgress(message="Loading example Data...", value=0.5, {
path <- "Xenium_V1_FF_Mouse_Brain_Coronal_Subset_CTX_HP_outs/"
xenium.obj <- LoadXenium(path, fov = "fov")
xenium.obj <- subset(xenium.obj, subset = nCount_Xenium > 0)
})
if (is.null(xenium.obj)){
v$scData_xenium <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(xenium.obj)
label_xenium <- "Example loaded"
updateActionButton(inputId = "loadexample_xenium", label = label_xenium)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData_xenium <- xenium.obj
})
}
})
observeEvent(input$load_xenium, {
withProgress(message="Loading and Processing Xenium Data...", value=0, {
exp.data_xenium <- LoadXenium(data.dir = parseDirPath(c(home = '.'), dir_xenium()), fov = "fov")
exp.data_xenium <- subset(exp.data_xenium, subset = nCount_Xenium > input$count_xenium)
incProgress(0.5, "Creating xenium object")
v$scData_xenium <- exp.data_xenium
print(v$scData_xenium)
shinyalert("Data uploaded", "Data uploaded, please click on Process button", type = "success", imageWidth = 10, imageHeight = 10)
})
}
)
observeEvent(input$reset_xenium, {
session$reload()
print("Reset done")
})
output$nFeature_xenium <- renderPlot({
if(is.null(v$scData_xenium)){
plotly_empty()
}else{
VlnPlot(v$scData_xenium, features = "nFeature_Xenium", pt.size = 0) + NoLegend()
}
})
output$nCount_xenium <- renderPlot({
if(is.null(v$scData_xenium)){
plotly_empty()
}else{
VlnPlot(v$scData_xenium, features = "nCount_Xenium", pt.size = 0) + NoLegend()
}
})
output$FeatureScatterPlot_xenium <- renderPlotly({
if(is.null(v$scData_xenium)){
plotly_empty()
}else{
print(FeatureScatter(v$scData_xenium, "nCount_Xenium", "nFeature_Xenium"))
}
})
output$xenium.gene.select <- renderUI({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
selectInput("xenium.gene", label = "Molecules to visualise",
choices = rownames(v$scData_xenium@assays$Xenium@counts), multiple = T, selected = rownames(v$scData_xenium@assays$Xenium@counts)[20])
}
})
output$markerplot_xenium <- renderPlot({
if(is.null(v$scData_xenium)){
plotly_empty()
}else{
ImageDimPlot(v$scData_xenium, fov = "fov", molecules = input$xenium.gene, nmols = 20000)
}
})
output$xenium.gene1.select <- renderUI({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
selectInput("xenium.gene1", label = "Molecules to visualise",
choices = rownames(v$scData_xenium@assays$Xenium@counts), selected = rownames(v$scData_xenium@assays$Xenium@counts)[20])
}
})
output$featureplot_xenium <- renderPlot({
if(is.null(v$scData_xenium)){
plotly_empty()
}else{
ImageFeaturePlot(v$scData_xenium, features = input$xenium.gene1, size = 0.75, cols = c("white", "red"))
}
})
observeEvent(input$crop_xenium, {
withProgress(message = "Cropping...", value = 0.3, {
cropped.coords <- Crop(v$scData_xenium[["fov"]], x = c(input$x1_xenium, input$x2_xenium), y = c(input$y1_xenium, input$y2_xenium), coords = "plot")
print(cropped.coords)
v$scData_xenium[["zoom"]] <- cropped.coords
# visualize cropped area with cell segmentations & selected molecules
DefaultBoundary(v$scData_xenium[["zoom"]]) <- "segmentation"
v$isCropXeniumdone <- TRUE
})
})
output$xenium.gene2.select <- renderUI({
if(is.null(v$scData_xenium)){
return(NULL)
}else{
selectInput("xenium.gene2", label = "Molecules to visualize",
choices = rownames(v$scData_xenium@assays$Xenium@counts), multiple = T, selected = rownames(v$scData_xenium@assays$Xenium@counts)[20])
}
})
output$zoom_xenium <- renderPlot({
if(is.null(v$scData_xenium) || is.null(v$isCropXeniumdone)){
plotly_empty()
}else{
ImageDimPlot(v$scData_xenium, fov = "zoom", axes = TRUE, border.color = "white", border.size = 0.1, cols = "polychrome", coord.fixed = FALSE, molecules = input$xenium.gene2, nmols = 10000)
}
})
output$gene_sp.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
if(input$deg_sp1 == "seurat_clusters"){
selectInput("gene_sp1", label = "Celltype1",
choices = as.vector(v$scData_spatial$seurat_clusters))
}
}
})
output$gene_sp1.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
if(input$deg_sp1 == "seurat_clusters"){
selectInput("gene_sp2", label = "Celltype2",
choices = as.vector(v$scData_spatial$seurat_clusters))
}
}
})
observeEvent(input$doVolcano_sp, {
tpmFiles <- v$scData_spatial
if (is.null(tpmFiles)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
DefaultAssay(v$scData_spatial) <- "SCT"
v$scData_spatial_subset <- subset(v$scData_spatial, subset = seurat_clusters == input$gene_sp1 | seurat_clusters == input$gene_sp2)
ips.markers_sp_a <- FindAllMarkers(v$scData_spatial_subset, only.pos = F, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1, assay = "SCT")
ips.markers_sp_b <- FindMarkers(v$scData_spatial_subset, ident.1 = input$gene_sp1, ident.2 = input$gene_sp2, only.pos = F, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$ips.markers_sp_a <- ips.markers_sp_a
v$ips.markers_sp_b <- ips.markers_sp_b
shinyalert("Pairwise DEGs done", "Pairwise DEGs done, please run GSEA Analysis", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$volcano.plot_sp1 <- renderPlot({
if(is.null(v$ips.markers_sp_b)){
return(NULL)
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
EnhancedVolcano(toptable = v$ips.markers_sp_b, lab = row.names(v$ips.markers_sp_b), x ="avg_log2FC", y ="p_val_adj", pointSize = 1, labSize = 5, legendLabSize = 12, axisLabSize = 12)
})
}
})
output$dega_sp.plot <- renderPlot({
if(is.null(v$scData_spatial_subset)){
return(NULL)
}else{
withProgress(message="Generating Volcano Plot...", value=0, {
v$ips.markers_sp_a %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData_spatial_subset, features = top10$gene, size = 5, angle = 45) + theme(axis.text.y = element_text(size = 8)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
observeEvent(input$doDeg_spatial, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
if(input$deg_spatial == "seurat_clusters"){
#DefaultAssay(v$scData_spatial) <- "SCT"
print(v$scData_spatial@meta.data)
v$scData_spatial$seurat_clusters -> Idents(v$scData_spatial)
ips.markers_spatial <- FindAllMarkers(v$scData_spatial, only.pos = FALSE, min.pct = input$min_pct_spatial, logfc.threshold = input$logfc_spatial, test.use = input$test.use_spatial, assay = "SCT")
v$ips.markers_spatial <- ips.markers_spatial
}
})
}
})
observeEvent(input$Vis_spatial, {
tpmFile_spatial <- v$scData_spatial
if (is.null(tpmFile_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Visualizing...", value=0, {
v$isVisSpatialdone <- TRUE
})
}
})
output$deg.gene.select_spatial <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
selectInput("deg.gene_spatial", label = "Gene to visualise",
choices = rownames(v$scData_spatial[["SCT"]]))
}
})
plotViolin_spatial <- reactive({
if(is.null(v$scData_spatial) || is.null(v$isVisSpatialdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
DefaultAssay(v$scData_spatial) <- "SCT"
VlnPlot(v$scData_spatial, input$deg.gene_spatial)
})
}
})
output$Deg.plot_spatial <- renderPlotly({
plotViolin_spatial()
})
output$download_violn_spatial <- downloadHandler(
filename = function(){"Violin plot (Spatial module).png"},
content = function(fname){
ggsave(fname,plotViolin_spatial(), height = 7, width = 7)
}
)
plotFeature_spatial <- reactive({
if(is.null(v$scData_spatial) || is.null(v$isVisSpatialdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
DefaultAssay(v$scData_spatial) <- "SCT"
SpatialFeaturePlot(v$scData_spatial, input$deg.gene_spatial)
})
}
})
output$Deg1.plot_spatial <- renderPlot({
plotFeature_spatial()
})
output$download_feature_spatial <- downloadHandler(
filename = function(){"Feature plot (Spatial module).png"},
content = function(fname){
ggsave(fname,plotFeature_spatial(), height = 7, width = 7)
}
)
plotRidge_spatial <- reactive({
if(is.null(v$scData_spatial) || is.null(v$isVisSpatialdone)){
plotly_empty()
}else{
withProgress(message="Visualizing...", value=0, {
DefaultAssay(v$scData_spatial) <- "SCT"
RidgePlot(v$scData_spatial, features = input$deg.gene_spatial)
})
}
})
output$Deg2.plot_spatial <- renderPlot({
plotRidge_spatial()
})
output$download_ridge_spatial <- downloadHandler(
filename = function(){"Ridge plot (Spatial module).png"},
content = function(fname){
ggsave(fname,plotRidge_spatial(), height = 7, width = 7)
}
)
output$Deg.heatmap_spatial <- renderPlot({
if(is.null(v$ips.markers_spatial)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$ips.markers_spatial %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$scData_spatial, features = top10$gene, size = 4, assay = "SCT") + theme(axis.text.y = element_text(size = 5)) + scale_fill_gradientn(colors = c("blue", "white", "red")) + NoLegend()
})
}
})
output$Deg.table_spatial <- DT::renderDataTable(
v$ips.markers_spatial, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$gsea.ct_sp1.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
selectInput("gsea_sp.ct1", label = "Celltype1",
choices = as.vector(v$scData_spatial$seurat_clusters))
}
})
output$gsea.ct_sp2.select <- renderUI({
if(is.null(v$scData_spatial)){
return(NULL)
}else{
selectInput("gsea_sp.ct2", label = "Celltype2",
choices = as.vector(v$scData_spatial$seurat_clusters))
}
})
observeEvent(input$gsea_sp, {
tpmFiles_spatial <- v$scData_spatial
if (is.null(tpmFiles_spatial)){
v$scData_spatial <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating gene set enrichment analysis...", value=0, {
if(input$species_gsea_sp == "Homo sapiens" & input$category_gsea_sp == "H"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_hs_go_sp <- msigdbr(species = "Homo sapiens", category = "H")
print(v$msigdbr_hs_go_sp)
v$pathways_sp <- split(x = v$msigdbr_hs_go_sp$gene_symbol, f = v$msigdbr_hs_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Mus musculus" & input$category_gsea_sp == "H"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_mm_go_sp <- msigdbr(species = "Mus musculus", category = "H")
print(v$msigdbr_mm_go_sp)
v$pathways_sp <- split(x = v$msigdbr_mm_go_sp$gene_symbol, f = v$msigdbr_mm_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Homo sapiens" & input$category_gsea_sp == "C2"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_hs_go_sp <- msigdbr(species = "Homo sapiens", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_hs_go_sp)
v$pathways_sp <- split(x = v$msigdbr_hs_go_sp$gene_symbol, f = v$msigdbr_hs_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Mus musculus" & input$category_gsea_sp == "C2"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_mm_go_sp <- msigdbr(species = "Mus musculus", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_mm_go_sp)
v$pathways_sp <- split(x = v$msigdbr_mm_go_sp$gene_symbol, f = v$msigdbr_mm_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Homo sapiens" & input$category_gsea_sp == "C5"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_hs_go_sp <- msigdbr(species = "Homo sapiens", category = "C5")
print(v$msigdbr_hs_go_sp)
v$pathways_sp <- split(x = v$msigdbr_hs_go_sp$gene_symbol, f = v$msigdbr_hs_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Mus musculus" & input$category_gsea_sp == "C5"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_mm_go_sp <- msigdbr(species = "Mus musculus", category = "C5")
print(v$msigdbr_mm_go_sp)
v$pathways_sp <- split(x = v$msigdbr_mm_go_sp$gene_symbol, f = v$msigdbr_mm_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Homo sapiens" & input$category_gsea_sp == "C7"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_hs_go_sp <- msigdbr(species = "Homo sapiens", category = "C7")
print(v$msigdbr_hs_go_sp)
v$pathways_sp <- split(x = v$msigdbr_hs_go_sp$gene_symbol, f = v$msigdbr_hs_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Mus musculus" & input$category_gsea_sp == "C7"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_mm_go_sp <- msigdbr(species = "Mus musculus", category = "C7")
print(v$msigdbr_mm_go_sp)
v$pathways_sp <- split(x = v$msigdbr_mm_go_sp$gene_symbol, f = v$msigdbr_mm_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Homo sapiens" & input$category_gsea_sp == "C8"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_hs_go_sp <- msigdbr(species = "Homo sapiens", category = "C8")
print(v$msigdbr_hs_go_sp)
v$pathways_sp <- split(x = v$msigdbr_hs_go_sp$gene_symbol, f = v$msigdbr_hs_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
if(input$species_gsea_sp == "Mus musculus" & input$category_gsea_sp == "C8"){
DefaultAssay(v$scData_spatial) <- "SCT"
v$msigdbr_mm_go_sp <- msigdbr(species = "Mus musculus", category = "C8")
print(v$msigdbr_mm_go_sp)
v$pathways_sp <- split(x = v$msigdbr_mm_go_sp$gene_symbol, f = v$msigdbr_mm_go_sp$gs_name)
print(v$pathways_sp)
v$markers_sp <- FindMarkers(v$scData_spatial, ident.1 = input$gsea_sp.ct1, ident.2 = input$gsea_sp.ct2, min.pct = input$min_pct_sp1, logfc.threshold = input$logfc_sp1, test.use = input$test.use_sp1)
v$markers_sp <- v$markers_sp %>% arrange(desc(avg_log2FC))
print(v$markers_sp)
v$markers_sp.log2FC <- v$markers_sp$avg_log2FC
names(v$markers_sp.log2FC) <- row.names(v$markers_sp)
v$markers_sp.log2FC <- sort(na.omit(v$markers_sp.log2FC), decreasing = TRUE)
print(v$markers_sp.log2FC)
v$fgseaRes_sp <- fgsea(pathways = v$pathways_sp, stats = v$markers_sp.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgseaRes_sp)
v$topPathwaysUp_sp <- v$fgseaRes_sp[ES > 0][head(order(pval), n=10), pathway]
v$topPathwaysDown_sp <- v$fgseaRes_sp[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_sp <- c(v$topPathwaysUp_sp, rev(v$topPathwaysDown_sp))
}
output$gsea_sp.done <- renderText(paste0("Gene set enrichment done!"))
v$isGSEAspdone <- TRUE
})
}
})
output$gsea_sp.select <- renderUI({
if(is.null(v$pathways_sp)){
return(NULL)
}else{
selectInput("gsea_sp.pathway", label = "Gene set to visualise",
choices = names(v$pathways_sp))
}
})
output$gsea_sp_plot <- renderPlot({
if(is.null(v$pathways_sp) || is.null(v$isGSEAspdone)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotEnrichment(v$pathways_sp[[input$gsea_sp.pathway]], v$markers_sp.log2FC) + labs(title=input$gsea_sp.pathway)
})
}
})
output$gsea_sp_plot1 <- renderPlot({
if(is.null(v$pathways_sp) || is.null(v$isGSEAspdone)){
return(NULL)
}else{
withProgress(message="Generating GSEA plot...", value=0, {
plotGseaTable(v$pathways_sp[v$topPathways_sp], v$markers_sp.log2FC, v$fgseaRes_sp, gseaParam=0.5)
})
}
})
output$gsea_sp.table <- DT::renderDataTable(
v$fgseaRes_sp, options = list(scrollX = TRUE, scrollY = "400px"), server = FALSE)
output$download_gsea_sp.table <- downloadHandler(
filename = function(){"GSEA Results.csv"},
content = function(fname){
withProgress(message="Downloading GSEA Results...", value=0, {
fwrite(v$fgseaRes_sp, fname)
})
}
)
#---------------------Single cell ATAC-seq-----------------##
output$atac_image <- renderImage({
list(src = "www/atac_fig.png",
width = 800,
height = 450)
}, deleteFile = FALSE)
observeEvent(input$loadButton_atac, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
meta_atac <- input$meta_atac
meta_atac <- v$meta_atac
print(tpmFiles_atac)
print(meta_atac)
names.field <- input$field
#tpmFiles_atac <- input$tpmFiles_atac
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(tpmFiles_atac$datapath)
print(tpmFiles_atac$name)
#print(file.exists(paste(tpmFiles_atac$datapath[1], "/", tpmFiles_atac$name[1], sep="")))
exp.data_atac <- Read10X_h5(tpmFiles_atac$datapath)
additional.ident <- NULL
if(!is.null(meta_atac)){
anno.data_atac <- read.csv(meta_atac$datapath[1], header = T, row.names = 1)
print(anno.data_atac)
#to.append <- apply(anno.data, 1, paste, collapse = "_")
#colnames(exp.data) <- to.append
#names.field <- match(input$groupby, colnames(anno.data))
#additional.ident <- data.frame(data.frame(anno.data[,-1], row.names = to.append))
#additional.ident[] <- lapply(additional.ident, factor)
}
incProgress(0.5, "Creating Chromatin Assay")
print(exp.data_atac)
#frag <- CreateFragmentObject(path = parseDirPath(c(home = '.'), dir_atac()))
#print(frag)
chrom_assay_atac <- CreateChromatinAssay(counts = exp.data_atac, sep = c(":", "-"), genome = 'hg19', fragments = as.character(parseFilePaths(c(home = '.'), dir_atac())$datapath), min.cells = input$min.cells_atac, min.features = input$min.genes_atac)
print (chrom_assay_atac)
incProgress(0.5, "Creating Seurat Object")
sObj_atac <- CreateSeuratObject(chrom_assay_atac,
assay = "peaks",
project = input$projName4,
meta.data = anno.data_atac)
print(sObj_atac)
print(sObj_atac[['peaks']])
print(granges(sObj_atac))
v$atacData <- sObj_atac
print(v$atacData)
shinyalert("Data uploaded", "Data uploaded", type = "success", imageWidth = 10, imageHeight = 10)
})
}
dir.create("Seurat_ATACseq_results")
})
observeEvent(input$loadexample_atac, {
withProgress(message="Loading example Data...", value=0.5, {
tpmFiles_atac <- Read10X_h5(filename = "atac_pbmc_500_v1/filtered_peak_bc_matrix.h5")
meta_atac <- read.csv('atac_pbmc_500_v1/singlecell.csv', header = T, row.names = 1)
chrom_assay_atac <- CreateChromatinAssay(counts = tpmFiles_atac, sep = c(":", "-"), genome = 'hg19', fragments = 'atac_pbmc_500_v1/fragments.tsv.gz', min.cells = 3, min.features = 200)
sObj_atac <- CreateSeuratObject(chrom_assay_atac, assay = "peaks", project = input$projName4, meta.data = meta_atac)
print(sObj_atac)
annotations_atac <- GetGRangesFromEnsDb(ensdb = EnsDb.Hsapiens.v86)
seqlevelsStyle(annotations_atac) <- 'UCSC'
genome(annotations_atac) <- "hg19"
print("test")
Annotation(sObj_atac) <- annotations_atac
sObj_atac <- NucleosomeSignal(object = sObj_atac)
print("test2")
sObj_atac <- TSSEnrichment(object = sObj_atac, fast = FALSE)
print("test1")
sObj_atac$pct_reads_in_peaks <- sObj_atac$peak_region_fragments / sObj_atac$passed_filters * 100
sObj_atac$blacklist_ratio <- sObj_atac$blacklist_region_fragments / sObj_atac$peak_region_fragments
sObj_atac$high.tss <- ifelse(sObj_atac$TSS.enrichment > 2, 'High', 'Low')
sObj_atac$nucleosome_group <- ifelse(sObj_atac$nucleosome_signal > 4, 'NS > 4', 'NS < 4')
v$tpmFiles_atac <- tpmFiles_atac
v$meta_atac <- meta_atac
v$atacData <- sObj_atac
})
label1 <- "Example loaded"
updateActionButton(inputId = "loadexample_atac", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
})
observeEvent(input$reset_atac, {
session$reload()
print("Reset done")
})
shinyFileChoose(
input,
'dir_atac',
roots = c(home = '.'),
filetypes = c('', 'txt', 'bigWig', "tsv", "csv", "bw", "gz", "tbi")
)
dir_atac <- reactive(input$dir_atac)
output$dir_atac <- renderPrint({ # use renderText instead of renderPrint
as.character(parseFilePaths(c(home = '.'), dir_atac())$datapath)
})
output$countdataDT_atac <- renderDataTable({
if(!is.null(v$atacData))
{
if(ncol(v$atacData) > 20 )
return(as.matrix(v$atacData@assays$peaks@counts[1:20,]))
}
}, server = FALSE)
observeEvent(input$create_seurat_atac, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Loading and Processing Data...", value=0, {
print(v$atacData)
chrom_assay_atac <- CreateChromatinAssay(counts = v$tpmFiles_atac, sep = c(":", "-"), genome = 'hg19', fragments = 'atac_pbmc_500_v1/fragments.tsv.gz', min.cells = input$min.cells_atac, min.features = input$min.genes_atac)
sObj_atac <- CreateSeuratObject(chrom_assay_atac, assay = "peaks", project = input$projName4, meta.data = v$meta_atac)
print(sObj_atac)
annotations_atac <- GetGRangesFromEnsDb(ensdb = EnsDb.Hsapiens.v86)
seqlevelsStyle(annotations_atac) <- 'UCSC'
genome(annotations_atac) <- "hg19"
print("test")
Annotation(sObj_atac) <- annotations_atac
sObj_atac <- NucleosomeSignal(object = sObj_atac)
print("test2")
sObj_atac <- TSSEnrichment(object = sObj_atac, fast = FALSE)
print("test1")
sObj_atac$pct_reads_in_peaks <- sObj_atac$peak_region_fragments / sObj_atac$passed_filters * 100
sObj_atac$blacklist_ratio <- sObj_atac$blacklist_region_fragments / sObj_atac$peak_region_fragments
sObj_atac$high.tss <- ifelse(sObj_atac$TSS.enrichment > 2, 'High', 'Low')
sObj_atac$nucleosome_group <- ifelse(sObj_atac$nucleosome_signal > 4, 'NS > 4', 'NS < 4')
v$atacData <- sObj_atac
shinyalert("Seurat object created", "Seurat object created", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$TSSPlot <- renderPlotly({
if(is.null(v$atacData)){
plotly_empty()
}else{
TSSPlot(v$atacData, group.by = 'high.tss')
}
})
output$FragmentHistogram <- renderPlotly({
if(is.null(v$atacData)){
plotly_empty()
}else{
FragmentHistogram(object = v$atacData, group.by = 'nucleosome_group')
}
})
output$Vlnplot_atac_1 <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
VlnPlot(object = v$atacData, features = 'pct_reads_in_peaks')
}
})
output$Vlnplot_atac_2 <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
VlnPlot(object = v$atacData, features = 'peak_region_fragments')
}
})
output$Vlnplot_atac_3 <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
VlnPlot(object = v$atacData, features = 'TSS.enrichment')
}
})
output$Vlnplot_atac_4 <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
VlnPlot(object = v$atacData, features = 'blacklist_ratio')
}
})
output$Vlnplot_atac_5 <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
VlnPlot(object = v$atacData, features = 'nucleosome_signal')
}
})
observeEvent(input$donorm_ATAC, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Performing Normalization...", value = 0,{
incProgress(0.5, message = "Running Dimension Reduction...")
v$atacData <- RunTFIDF(v$atacData)
v$atacData <- FindTopFeatures(v$atacData, min.cutoff = input$top_features_atac)
v$atacData <- RunSVD(v$atacData, n = input$npc_atac)
all_atac.genes <- rownames(v$atacData)
v$atacData <- ScaleData(v$atacData, features = all_atac.genes)
output$normalize_atac.done <- renderText(paste0("Normalization done!"))
v$isNormalizeATACdone <- TRUE
})
}
})
output$DepthCor_ATAC <- renderPlotly({
if(is.null(v$isNormalizeATACdone)){
plotly_empty()
}else{
DepthCor(v$atacData, n = input$npc_atac)
}
})
##---------------Clustering of ATAC-seq data-------------------
observeEvent(input$doCluster_ATAC, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Performing clustering...", value = 0.3,{
incProgress(0.5, message = "Running Clustering...")
v$atacData <- FindNeighbors(object = v$atacData, reduction = 'lsi', dims = 2:input$dim.used_atac)
v$atacData <- FindClusters(object = v$atacData, resolution = input$clus.res_atac, verbose = FALSE, algorithm = 3)
output$cluster_atac.done <- renderText(paste0("Clustering done!"))
v$isClusterATACdone <- TRUE
})
}
})
output$cluster_ATAC <- renderPlotly({
if(is.null(v$isClusterATACdone)){
plotly_empty()
}else{
DimPlot(object = v$atacData, reduction = "umap", label = TRUE)
}
})
##---------------UMAP on ATAC-seq data-------------------
observeEvent(input$doUMAP_ATAC, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running UMAP...", value = 0.3,{
incProgress(0.5, message = "Running UMAP...")
v$atacData <- RunUMAP(object = v$atacData, reduction = 'lsi', dims = 2:input$dim.used_atac)
output$umap_atac.done <- renderText(paste0("UMAP done!"))
v$isUMAPATACdone <- TRUE
})
}
})
output$Umap_ATAC <- renderPlotly({
if(is.null(v$isUMAPATACdone)){
plotly_empty()
}else{
DimPlot(object = v$atacData, label = TRUE)
}
})
##---------------TSNE on ATAC-seq data-------------------
observeEvent(input$doTSNE_ATAC, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message = "Running TSNE...", value = 0.3,{
incProgress(0.5, message = "Running TSNE...")
v$atacData <- RunTSNE(object = v$atacData, reduction = 'lsi', dims = 2:input$dim.used_atac)
output$tsne_atac.done <- renderText(paste0("TSNE done!"))
v$isTSNEATACdone <- TRUE
})
}
})
output$Tsne_ATAC <- renderPlotly({
if(is.null(v$isTSNEATACdone)){
plotly_empty()
}else{
DimPlot(object = v$atacData, reduction = "tsne", label = TRUE)
}
})
observeEvent(input$calc_gene_activity_atac, {
#tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Calculating gene activity...", value=0, {
gene.activities <- GeneActivity(v$atacData, assay = 'peaks')
v$atacData[['RNA']] <- CreateAssayObject(counts = gene.activities)
v$atacData <- NormalizeData(
object = v$atacData,
assay = 'RNA',
normalization.method = 'LogNormalize',
scale.factor = median(v$atacData$nCount_RNA)
)
DefaultAssay(v$atacData) <- 'RNA'
v$gene.activities <- gene.activities
v$isGeneActivitydone <- TRUE
})
}
})
output$gene_activity.select <- renderUI({
if(is.null(v$atacData) || is.null(v$isGeneActivitydone)){
return(NULL)
}else{
selectInput("gene_activity", label = "Gene activity to visualise",
choices = rownames(v$atacData), selected = rownames(v$atacData)[10])
}
})
plotGeneActivity <- reactive({
if(is.null(v$atacData) || is.null(v$isGeneActivitydone)){
plotly_empty()
}else{
withProgress(message="Generating Gene Activity Plot...", value=0, {
FeaturePlot(object = v$atacData,
features = input$gene_activity,
pt.size = 0.1
)
})
}
})
output$gene_activity.plot <- renderPlotly({
plotGeneActivity()
})
output$download_gene_activity <- downloadHandler(
filename = function(){"Gene Activity plot.png"},
content = function(fname){
ggsave(fname,plotGeneActivity(), height = 7, width = 7)
}
)
observeEvent(input$load_eg_scRNA, {
withProgress(message="Loading example reference scRNA-seq data...", value=0.5, {
tpmFiles_ex_scRNA <- read.table('scRNA/pbmc2.txt', header = T, row.names = 1, check.names = F)
tpmFiles_ex_scRNA <- CreateSeuratObject(tpmFiles_ex_scRNA)
tpmFiles_ex_scRNA[["percent.mt"]] <- PercentageFeatureSet(tpmFiles_ex_scRNA, pattern = "^MT-")
tpmFiles_ex_scRNA <- NormalizeData(tpmFiles_ex_scRNA)
tpmFiles_ex_scRNA <- FindVariableFeatures(tpmFiles_ex_scRNA, selection.method = "vst", nfeatures = 2000)
all.genes <- rownames(tpmFiles_ex_scRNA)
tpmFiles_ex_scRNA <- ScaleData(tpmFiles_ex_scRNA, features = all.genes)
tpmFiles_ex_scRNA <- RunPCA(tpmFiles_ex_scRNA, features = VariableFeatures(object = tpmFiles_ex_scRNA))
tpmFiles_ex_scRNA <- FindNeighbors(tpmFiles_ex_scRNA, dims = 1:30)
tpmFiles_ex_scRNA <- FindClusters(tpmFiles_ex_scRNA, resolution = 1)
tpmFiles_ex_scRNA <- RunUMAP(tpmFiles_ex_scRNA, dims = 1:30)
ref = readRDS('ref.rds')
tpmFiles_ex_scRNA.data.average = AverageExpression(tpmFiles_ex_scRNA)
tpmFiles_ex_scRNA.data.average = round(tpmFiles_ex_scRNA.data.average$RNA, 2)
if(input$cellatlas_atac == "all"){
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, ref)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, adipose1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, adrenal_gland1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, blood1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, bone_marrow1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, brain1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, breast1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, breast_milk1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, eye1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, gut1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, heart1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, kidney1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, liver1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, lung1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, pancreas1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, PDAC1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, skin1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, testis1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, thymus1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
if(input$cellatlas_atac == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
res = FastIntegration::CELLiD(tpmFiles_ex_scRNA.data.average, tonsil1)
print(res)
tpmFiles_ex_scRNA$primary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),1]
tpmFiles_ex_scRNA$secondary.predict = res[as.numeric(tpmFiles_ex_scRNA$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_scRNA@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
v$isCELLAtacdone <- TRUE
}
})
if (is.null(tpmFiles_ex_scRNA)){
v$scData1 <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "load_eg_scRNA", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData1 <- tpmFiles_ex_scRNA
v$isCELLAtacdone <- TRUE
#v$transfer.anchors <- FindTransferAnchors(
# reference = tpmFiles_ex_scRNA,
# query = v$atacData,
# reduction = 'cca'
#)
#v$predicted.labels <- TransferData(
# anchorset = v$transfer.anchors,
# refdata = tpmFiles_ex_scRNA$primary.predict,
# weight.reduction = v$atacData[['lsi']],
# dims = 2:30
#)
#print(v$predicted.labels)
#v$atacData <- AddMetaData(object = v$atacData, metadata = v$predicted.labels)
#print(v$atacData@meta.data)
})
}
})
observe({
if(input$load_user_scRNA > 0){
print('2')
session$sendCustomMessage("myCallbackHandler4", "2")
}
})
#observeEvent(input$doct_atac, {
# if(is.null(v$atacData)){
# return(NULL)
# }else{
# withProgress(message="Annotating celltypes for scATAC-seq data...", value=0, {
# v$transfer.anchors <- FindTransferAnchors(
# reference = v$scData1,
# query = v$atacData,
# reduction = 'cca'
# )
# v$predicted.labels <- TransferData(
# anchorset = v$transfer.anchors,
# refdata = v$scData1$primary.predict,
# weight.reduction = v$atacData[['lsi']],
# dims = 2:30
# )
# v$atacData <- AddMetaData(object = v$atacData, metadata = v$predicted.labels)
# })
# }
#})
output$cellanno_atac.plot <- renderPlot({
if(is.null(v$scData1) || is.null(v$isCELLAtacdone)){
plotly_empty()
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$scData1, group.by = 'primary.predict', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scRNA-seq')
})
}
})
plotCellAnnoAtac <- reactive({
if(is.null(v$atacData) || is.null(v$isCELLAtacsdone)){
return(NULL)
}else{
withProgress(message="Generating Plot...", value=0, {
DimPlot(object = v$atacData, group.by = 'predicted.id', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scATAC-seq')
})
}
})
output$cellanno_atac1.plot <- renderPlot({
plotCellAnnoAtac()
})
output$download_cellanno_plot <- downloadHandler(
filename = function(){"Cell annotation plot.png"},
content = function(fname){
ggsave(fname,plotCellAnnoAtac(), height = 7, width = 7)
}
)
#output$cellanno_atac1.plot <- renderPlot({
# if(is.null(v$atacData) || is.null(v$isCELLAtacdone)){
# plotly_empty()
# }else{
# withProgress(message="Generating DEG Plot...", value=0, {
# DimPlot(object = v$atacData, group.by = 'predicted.id', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scATAC-seq')
# })
# }
#})
observeEvent(input$load_eg1_scRNA, {
withProgress(message="Loading example reference scRNA-seq data...", value=0.5, {
tpmFiles_ex_scRNA1 <- read.table('scRNA/pbmc.txt', header = T, row.names = 1, check.names = F)
tpmFiles_ex_scRNA1 <- CreateSeuratObject(tpmFiles_ex_scRNA1)
tpmFiles_ex_scRNA1[["percent.mt"]] <- PercentageFeatureSet(tpmFiles_ex_scRNA1, pattern = "^MT-")
tpmFiles_ex_scRNA1 <- NormalizeData(tpmFiles_ex_scRNA1)
tpmFiles_ex_scRNA1 <- FindVariableFeatures(tpmFiles_ex_scRNA1, selection.method = "vst", nfeatures = 2000)
all.genes <- rownames(tpmFiles_ex_scRNA1)
tpmFiles_ex_scRNA1 <- ScaleData(tpmFiles_ex_scRNA1, features = all.genes)
tpmFiles_ex_scRNA1 <- RunPCA(tpmFiles_ex_scRNA1, features = VariableFeatures(object = tpmFiles_ex_scRNA1))
tpmFiles_ex_scRNA1 <- FindNeighbors(tpmFiles_ex_scRNA1, dims = 1:30)
tpmFiles_ex_scRNA1 <- FindClusters(tpmFiles_ex_scRNA1, resolution = 1)
tpmFiles_ex_scRNA1 <- RunUMAP(tpmFiles_ex_scRNA1, dims = 1:30)
sc <- reticulate::import("scanpy", convert = FALSE)
ct <- reticulate::import("celltypist", convert = FALSE)
sceasy::convertFormat(tpmFiles_ex_scRNA1, from = "seurat", to = "anndata", outFile = 'ct_scrna.h5ad')
v$adata = sc$read_h5ad('ct_scrna.h5ad')
v$res = ct$annotate(filename = 'ct_scrna.h5ad', model = input$celltypistatlas_atac, majority_voting=T)
print(v$res)
v$adata = v$res$to_adata()
print("fff")
v$adata$obs$to_csv('celltypist_predict.csv')
v$meta_1 <- read.csv('celltypist_predict.csv', header = T, row.names = 1)
tpmFiles_ex_scRNA1 <- AddMetaData(tpmFiles_ex_scRNA1, metadata = v$meta_1)
tpmFiles_ex_scRNA1$primary.predict <- tpmFiles_ex_scRNA1$majority_voting
tpmFiles_ex_scRNA1$secondary.predict <- tpmFiles_ex_scRNA1$predicted_labels
print(tpmFiles_ex_scRNA1@meta.data)
})
if (is.null(tpmFiles_ex_scRNA1)){
v$scData1 <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "load_eg1_scRNA", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData1 <- tpmFiles_ex_scRNA1
v$isCELLAtac1done <- TRUE
})
}
})
output$cellanno1_atac.plot <- renderPlot({
if(is.null(v$scData1) || is.null(v$isCELLAtac1done)){
plotly_empty()
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$scData1, group.by = 'primary.predict', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scRNA-seq')
})
}
})
plotCellAnnoAtac1 <- reactive({
if(is.null(v$atacData) || is.null(v$isCELLAtac1done)){
plotly_empty()
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$atacData, group.by = 'predicted.id', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scATAC-seq')
})
}
})
output$cellanno1_atac1.plot <- renderPlotly({
plotCellAnnoAtac1()
})
output$download_cellanno_plot <- downloadHandler(
filename = function(){"Cell annotation plot.png"},
content = function(fname){
ggsave(fname,plotCellAnnoAtac1(), height = 7, width = 7)
}
)
observe({
if(input$load_user1_scRNA > 0){
print('2')
session$sendCustomMessage("myCallbackHandler4a", "2")
}
})
observeEvent(input$load_eg_scRNA_a, {
withProgress(message="Loading example reference scRNA-seq data...", value=0.5, {
tpmFiles_ex_Intg <- read.table("integration/concatenated_expr_data.txt", header = T, row.names = 1, check.names = F, sep = '\t')
#scH5 <- input$scH5
annoFile_ex_Intg <- read.table("integration/metadata.txt", header = T, row.names = 1, sep = '\t')
#tpmFiles_ex_Intg <- read.table('scRNA/pbmc.txt', header = T, row.names = 1, check.names = F)
tpmFiles_ex_Intg <- CreateSeuratObject(counts = tpmFiles_ex_Intg, meta.data = annoFile_ex_Intg)
tpmFiles_ex_Intg[["percent.mt"]] <- PercentageFeatureSet(tpmFiles_ex_Intg, pattern = "^MT-")
tpmFiles_ex_Intg.list <- SplitObject(tpmFiles_ex_Intg, split.by = "batch")
for (i in 1:length(tpmFiles_ex_Intg.list)) {
tpmFiles_ex_Intg.list[[i]] <- NormalizeData(tpmFiles_ex_Intg.list[[i]], verbose = FALSE)
tpmFiles_ex_Intg.list[[i]] <- FindVariableFeatures(tpmFiles_ex_Intg.list[[i]], selection.method = "vst", nfeatures = 2000, verbose = FALSE)
}
print(tpmFiles_ex_Intg)
print(tpmFiles_ex_Intg.list)
features <- SelectIntegrationFeatures(object.list = tpmFiles_ex_Intg.list, nfeatures = 2000)
print(features)
tpmFiles_ex_Intg.anchors <- FindIntegrationAnchors(object.list = tpmFiles_ex_Intg.list, anchor.features = features)
tpmFiles_ex_Intg <- IntegrateData(anchorset = tpmFiles_ex_Intg.anchors)
print(tpmFiles_ex_Intg.anchors)
print(tpmFiles_ex_Intg)
DefaultAssay(tpmFiles_ex_Intg) <- "integrated"
tpmFiles_ex_Intg <- ScaleData(tpmFiles_ex_Intg, verbose = FALSE)
tpmFiles_ex_Intg <- RunPCA(tpmFiles_ex_Intg, verbose = FALSE)
tpmFiles_ex_Intg <- FindNeighbors(tpmFiles_ex_Intg, reduction = "pca", dims = 1:30)
tpmFiles_ex_Intg <- FindClusters(tpmFiles_ex_Intg, resolution = 1)
tpmFiles_ex_Intg <- RunUMAP(tpmFiles_ex_Intg, reduction = "pca", dims = 1:30, spread = 1)
print(tpmFiles_ex_Intg)
ref = readRDS('ref.rds')
tpmFiles_ex_Intg.data.average = AverageExpression(tpmFiles_ex_Intg)
tpmFiles_ex_Intg.data.average = round(tpmFiles_ex_Intg.data.average$RNA, 2)
if(input$cellatlas_atac_a == "all"){
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, ref)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "adipose"){
adipose <- colnames(ref)[grepl("adipose",colnames(ref))]
adipose1 <- ref[,adipose]
colnames(adipose1) <- gsub("--adipose","",colnames(adipose1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, adipose1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "adrenal_gland"){
adrenal_gland <- colnames(ref)[grepl("adrenal_gland",colnames(ref))]
adrenal_gland1 <- ref[,adrenal_gland]
colnames(adrenal_gland1) <- gsub("--adrenal_gland","",colnames(adrenal_gland1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, adrenal_gland1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "blood"){
blood <- colnames(ref)[grepl("blood",colnames(ref))]
blood1 <- ref[,blood]
colnames(blood1) <- gsub("--blood","",colnames(blood1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, blood1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "bone_marrow"){
bone_marrow <- colnames(ref)[grepl("bone_marrow",colnames(ref))]
bone_marrow1 <- ref[,bone_marrow]
colnames(bone_marrow1) <- gsub("--bone_marrow","",colnames(bone_marrow1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, bone_marrow1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "brain"){
brain <- colnames(ref)[grepl("brain",colnames(ref))]
brain1 <- ref[,brain]
colnames(brain1) <- gsub("--brain","",colnames(brain1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, brain1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "breast"){
breast <- colnames(ref)[grepl("breast",colnames(ref))]
breast1 <- ref[,breast]
colnames(breast1) <- gsub("--breast","",colnames(breast1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, breast1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "breast_milk"){
breast_milk <- colnames(ref)[grepl("breast_milk",colnames(ref))]
breast_milk1 <- ref[,breast_milk]
colnames(breast_milk1) <- gsub("--breast_milk","",colnames(breast_milk1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, breast_milk1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "eye"){
eye <- colnames(ref)[grepl("eye",colnames(ref))]
eye1 <- ref[,eye]
colnames(eye1) <- gsub("--eye","",colnames(eye1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, eye1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "gut"){
gut <- colnames(ref)[grepl("gut",colnames(ref))]
gut1 <- ref[,gut]
colnames(gut1) <- gsub("--gut","",colnames(gut1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, gut1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "heart"){
heart <- colnames(ref)[grepl("heart",colnames(ref))]
heart1 <- ref[,heart]
colnames(heart1) <- gsub("--heart","",colnames(heart1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, heart1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "kidney"){
kidney <- colnames(ref)[grepl("kidney",colnames(ref))]
kidney1 <- ref[,kidney]
colnames(kidney1) <- gsub("--kidney","",colnames(kidney1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, kidney1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "liver"){
liver <- colnames(ref)[grepl("liver",colnames(ref))]
liver1 <- ref[,liver]
colnames(liver1) <- gsub("--liver","",colnames(liver1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, liver1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "lung"){
lung <- colnames(ref)[grepl("lung",colnames(ref))]
lung1 <- ref[,lung]
colnames(lung1) <- gsub("--lung","",colnames(lung1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, lung1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "pancreas"){
pancreas <- colnames(ref)[grepl("pancreas",colnames(ref))]
pancreas1 <- ref[,pancreas]
colnames(pancreas1) <- gsub("--pancreas","",colnames(pancreas1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, pancreas1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "PDAC"){
PDAC <- colnames(ref)[grepl("PDAC",colnames(ref))]
PDAC1 <- ref[,PDAC]
colnames(PDAC1) <- gsub("--PDAC","",colnames(PDAC1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, PDAC1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "skin"){
skin <- colnames(ref)[grepl("skin",colnames(ref))]
skin1 <- ref[,skin]
colnames(skin1) <- gsub("--skin","",colnames(skin1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, skin1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "testis"){
testis <- colnames(ref)[grepl("testis",colnames(ref))]
testis1 <- ref[,testis]
colnames(testis1) <- gsub("--testis","",colnames(testis1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, testis1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "thymus"){
thymus <- colnames(ref)[grepl("thymus",colnames(ref))]
thymus1 <- ref[,thymus]
colnames(thymus1) <- gsub("--thymus","",colnames(thymus1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, thymus1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
if(input$cellatlas_atac_a == "tonsil"){
tonsil <- colnames(ref)[grepl("tonsil",colnames(ref))]
tonsil1 <- ref[,tonsil]
colnames(tonsil1) <- gsub("--tonsil","",colnames(tonsil1))
res = FastIntegration::CELLiD(tpmFiles_ex_Intg.data.average, tonsil1)
print(res)
tpmFiles_ex_Intg$primary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),1]
tpmFiles_ex_Intg$secondary.predict = res[as.numeric(tpmFiles_ex_Intg$seurat_clusters),2]
newheaders <- c("primary.predict", "secondary.predict", "primary score", "secondary score")
colnames(res) <- newheaders
print(tpmFiles_ex_Intg@meta.data)
output$CELLiD.done <- renderText(paste0("Cell type identification done!"))
write.table(res, "output_CELLiD.txt", sep = "\t", quote = F, row.names = F, col.names = F)
}
})
if (is.null(tpmFiles_ex_Intg)){
v$scData1 <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "load_eg_scRNA_a", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData1 <- tpmFiles_ex_Intg
})
}
})
observe({
if(input$load_user_scRNA_a > 0){
print('2')
session$sendCustomMessage("myCallbackHandler_4", "2")
}
})
output$cellanno_atac.plot_a <- renderPlot({
if(is.null(v$scData1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$scData1, group.by = 'primary.predict', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scRNA-seq')
})
}
})
output$cellanno_atac1.plot_a <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$atacData, group.by = 'predicted.id', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scATAC-seq')
})
}
})
observeEvent(input$load_eg1_scRNA_a, {
withProgress(message="Loading example reference scRNA-seq data...", value=0.5, {
tpmFiles_ex_Intg <- read.table("integration/concatenated_expr_data.txt", header = T, row.names = 1, check.names = F, sep = '\t')
#scH5 <- input$scH5
annoFile_ex_Intg <- read.table("integration/metadata.txt", header = T, row.names = 1, sep = '\t')
#tpmFiles_ex_Intg <- read.table('scRNA/pbmc.txt', header = T, row.names = 1, check.names = F)
tpmFiles_ex_Intg <- CreateSeuratObject(counts = tpmFiles_ex_Intg, meta.data = annoFile_ex_Intg)
tpmFiles_ex_Intg[["percent.mt"]] <- PercentageFeatureSet(tpmFiles_ex_Intg, pattern = "^MT-")
tpmFiles_ex_Intg.list <- SplitObject(tpmFiles_ex_Intg, split.by = "batch")
for (i in 1:length(tpmFiles_ex_Intg.list)) {
tpmFiles_ex_Intg.list[[i]] <- NormalizeData(tpmFiles_ex_Intg.list[[i]], verbose = FALSE)
tpmFiles_ex_Intg.list[[i]] <- FindVariableFeatures(tpmFiles_ex_Intg.list[[i]], selection.method = "vst", nfeatures = 2000, verbose = FALSE)
}
print(tpmFiles_ex_Intg)
print(tpmFiles_ex_Intg.list)
features <- SelectIntegrationFeatures(object.list = tpmFiles_ex_Intg.list, nfeatures = 2000)
print(features)
tpmFiles_ex_Intg.anchors <- FindIntegrationAnchors(object.list = tpmFiles_ex_Intg.list, anchor.features = features)
tpmFiles_ex_Intg <- IntegrateData(anchorset = tpmFiles_ex_Intg.anchors)
print(tpmFiles_ex_Intg.anchors)
print(tpmFiles_ex_Intg)
DefaultAssay(tpmFiles_ex_Intg) <- "integrated"
tpmFiles_ex_Intg <- ScaleData(tpmFiles_ex_Intg, verbose = FALSE)
tpmFiles_ex_Intg <- RunPCA(tpmFiles_ex_Intg, verbose = FALSE)
tpmFiles_ex_Intg <- FindNeighbors(tpmFiles_ex_Intg, reduction = "pca", dims = 1:30)
tpmFiles_ex_Intg <- FindClusters(tpmFiles_ex_Intg, resolution = 1)
tpmFiles_ex_Intg <- RunUMAP(tpmFiles_ex_Intg, reduction = "pca", dims = 1:30, spread = 1)
print(tpmFiles_ex_Intg)
sc <- reticulate::import("scanpy", convert = FALSE)
ct <- reticulate::import("celltypist", convert = FALSE)
sceasy::convertFormat(tpmFiles_ex_Intg, from = "seurat", to = "anndata", outFile = 'ct_intg.h5ad')
v$adata = sc$read_h5ad('ct_intg.h5ad')
v$res1 = ct$annotate(filename = 'ct_intg.h5ad', model = input$celltypistatlas_atac_a, majority_voting=T)
print(v$res1)
v$adata = v$res1$to_adata()
print("fff")
v$adata$obs$to_csv('celltypist_predict.csv')
v$meta_1a <- read.csv('celltypist_predict.csv', header = T, row.names = 1)
tpmFiles_ex_Intg <- AddMetaData(tpmFiles_ex_Intg, metadata = v$meta_1a)
tpmFiles_ex_Intg$primary.predict <- tpmFiles_ex_Intg$majority_voting
tpmFiles_ex_Intg$secondary.predict <- tpmFiles_ex_Intg$predicted_labels
print(tpmFiles_ex_Intg@meta.data)
})
if (is.null(tpmFiles_ex_Intg)){
v$scData1 <- NULL
}else{
withProgress(message="Loading and Processing Data...", value=0.5, {
#print(file.exists(paste(tpmFiles_demo$datapath[1], "/", tpmFiles_demo$name[1], sep="")))
label1 <- "Example loaded"
updateActionButton(inputId = "load_eg1_scRNA_a", label = label1)
shinyalert("Loaded", "Example loaded.", type = "success", imageWidth = 10, imageHeight = 10)
v$scData1 <- tpmFiles_ex_Intg
})
}
})
observe({
if(input$load_user1_scRNA_a > 0){
print('2')
session$sendCustomMessage("myCallbackHandler_4a", "2")
}
})
output$cellanno1_atac.plot_a <- renderPlot({
if(is.null(v$scData1)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$scData1, group.by = 'primary.predict', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scRNA-seq')
})
}
})
output$cellanno1_atac1.plot_a <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DimPlot(object = v$atacData, group.by = 'predicted.id', label = TRUE, repel = TRUE) + NoLegend() + ggtitle('scATAC-seq')
})
}
})
observeEvent(input$doDeg_atac, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DE Peaks...", value=0, {
DefaultAssay(v$atacData) <- 'peaks'
print(v$atacData)
atac.markers <- FindAllMarkers(v$atacData, only.pos = FALSE, min.pct = input$min_pct_atac, logfc.threshold = input$logfc_atac, test.use = input$test.use_atac)
v$atac.markers <- atac.markers
})
}
})
output$Deg_atac.table <- DT::renderDataTable(
v$atac.markers, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$Deg_atac.plot <- renderPlotly({
if(is.null(v$atac.markers)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
v$atac.markers %>% group_by(cluster) %>% top_n(n = 10, wt = avg_log2FC) -> top10
DoHeatmap(v$atacData, features = top10$gene, size = 4, slot = "scale.data",angle = 30) + theme(axis.text.y = element_text(size = 4)) + NoLegend()
})
}
})
observeEvent(input$Vis_atac, {
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Visualizing...", value=0, {
v$isVisATACdone <- TRUE
})
}
})
output$deg.atac.select <- renderUI({
if(is.null(v$atac.markers)){
return(NULL)
}else{
selectInput("deg.atac", label = "Peaks to visualise",
choices = rownames(v$atac.markers))
}
})
plotViolinAtac <- reactive({
if(is.null(v$atac.markers) || is.null(v$isVisATACdone)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
VlnPlot(v$atacData, input$deg.atac)
})
}
})
output$Deg_atac1.plot <- renderPlotly({
plotViolinAtac()
})
output$download_violn_atac <- downloadHandler(
filename = function(){"Violin plot for ATAC.png"},
content = function(fname){
ggsave(fname,plotViolinAtac(), height = 7, width = 7)
}
)
plotFeatureAtac <- reactive({
if(is.null(v$atac.markers) || is.null(v$isVisATACdone)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(v$atacData, input$deg.atac)
})
}
})
output$Deg_atac2.plot <- renderPlotly({
plotFeatureAtac()
})
output$download_feature_atac <- downloadHandler(
filename = function(){"Feature plot for ATAC.png"},
content = function(fname){
ggsave(fname,plotFeatureAtac(), height = 7, width = 7)
}
)
plotRidgeAtac <- reactive({
if(is.null(v$atac.markers) || is.null(v$isVisATACdone)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
RidgePlot(v$atacData, features = input$deg.atac)
})
}
})
output$Deg_atac3.plot <- renderPlot({
plotRidgeAtac()
})
output$download_ridge_atac <- downloadHandler(
filename = function(){"Ridge plot for ATAC.png"},
content = function(fname){
ggsave(fname,plotRidgeAtac(), height = 7, width = 7)
}
)
output$peak_gene.atac.select <- renderUI({
if(is.null(v$atacData)){
return(NULL)
}else{
selectInput("gene_peak", label = "Genes to visualise",
choices = rownames(v$atacData@assays$RNA@counts), multiple = T, selected = rownames(v$atacData@assays$RNA@counts)[1])
}
})
plotCoverageAtac <- reactive({
if(is.null(v$atacData)){
plotly_empty()
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DefaultAssay(v$atacData) <- 'peaks'
CoveragePlot(
object = v$atacData,
region = input$gene_peak,
features = input$gene_peak,
extend.upstream = 500,
extend.downstream = 10000
)
})
}
})
output$coverage1.plot <- renderPlot({
plotCoverageAtac()
})
observeEvent(input$link_peak_genes, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Linking peaks to genes...", value=0, {
DefaultAssay(v$atacData) <- 'peaks'
# first compute the GC content for each peak
v$atacData <- RegionStats(v$atacData, genome = BSgenome.Hsapiens.UCSC.hg19)
# link peaks to genes
v$atacData <- LinkPeaks(
object = v$atacData,
peak.assay = "peaks",
expression.assay = "RNA",
genes.use = input$gene_peak)
v$peak <- as.data.frame(v$atacData@assays$peaks@links)
v$isLinkPeakdone <- TRUE
})
}
})
output$download_coverage_atac <- downloadHandler(
filename = function(){"Coverage plot for ATAC.png"},
content = function(fname){
ggsave(fname,plotCoverageAtac(), height = 7, width = 7)
}
)
output$coverage2.plot <- renderPlot({
if(is.null(v$atacData) || is.null(v$isLinkPeakdone)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DefaultAssay(v$atacData) <- 'peaks'
CoveragePlot(
object = v$atacData,
region = input$gene_peak,
features = input$gene_peak,
extend.upstream = 500,
extend.downstream = 10000
)
})
}
})
output$peaks.table <- DT::renderDataTable(
v$peak, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
##---------------GSEA of scATAC-seq-------------------##
output$gsea_atac.ct1.select <- renderUI({
if(is.null(v$atac.markers) || is.null(v$isCELLAtacsdone)){
return(NULL)
}else{
selectInput("gsea_atac.ct1", label = "Celltype1",
choices = as.vector(v$atacData$predicted.id))
}
})
output$gsea_atac.ct2.select <- renderUI({
if(is.null(v$atac.markers) || is.null(v$isCELLAtacsdone)){
return(NULL)
}else{
selectInput("gsea_atac.ct2", label = "Celltype2",
choices = as.vector(v$atacData$predicted.id))
}
})
observeEvent(input$gsea_atac, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating gene set enrichment analysis...", value=0, {
if(input$species_gsea_atac == "Homo sapiens" & input$category_gsea_atac == "H"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_hs_go_atac <- msigdbr(species = "Homo sapiens", category = "H")
print(v$msigdbr_hs_go_atac)
v$pathways_atac <- split(x = v$msigdbr_hs_go_atac$gene_symbol, f = v$msigdbr_hs_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Mus musculus" & input$category_gsea_atac == "H"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_mm_go_atac <- msigdbr(species = "Mus musculus", category = "H")
print(v$msigdbr_mm_go_atac)
v$pathways_atac <- split(x = v$msigdbr_mm_go_atac$gene_symbol, f = v$msigdbr_mm_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Homo sapiens" & input$category_gsea_atac == "C2"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_hs_go_atac <- msigdbr(species = "Homo sapiens", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_hs_go_atac)
v$pathways_atac <- split(x = v$msigdbr_hs_go_atac$gene_symbol, f = v$msigdbr_hs_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Mus musculus" & input$category_gsea_atac == "C2"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_mm_go_atac <- msigdbr(species = "Mus musculus", category = "C2") %>% dplyr::filter(gs_subcat %in% c("CP:KEGG", "CP:REACTOME"))
print(v$msigdbr_mm_go_atac)
v$pathways_atac <- split(x = v$msigdbr_mm_go_atac$gene_symbol, f = v$msigdbr_mm_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Homo sapiens" & input$category_gsea_atac == "C5"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_hs_go_atac <- msigdbr(species = "Homo sapiens", category = "C5")
print(v$msigdbr_hs_go_atac)
v$pathways_atac <- split(x = v$msigdbr_hs_go_atac$gene_symbol, f = v$msigdbr_hs_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Mus musculus" & input$category_gsea_atac == "C5"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_mm_go_atac <- msigdbr(species = "Mus musculus", category = "C5")
print(v$msigdbr_mm_go_atac)
v$pathways_atac <- split(x = v$msigdbr_mm_go_atac$gene_symbol, f = v$msigdbr_mm_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Homo sapiens" & input$category_gsea_atac == "C7"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_hs_go_atac <- msigdbr(species = "Homo sapiens", category = "C7")
print(v$msigdbr_hs_go_atac)
v$pathways_atac <- split(x = v$msigdbr_hs_go_atac$gene_symbol, f = v$msigdbr_hs_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Mus musculus" & input$category_gsea_atac == "C7"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_mm_go_atac <- msigdbr(species = "Mus musculus", category = "C7")
print(v$msigdbr_mm_go_atac)
v$pathways_atac <- split(x = v$msigdbr_mm_go_atac$gene_symbol, f = v$msigdbr_mm_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Homo sapiens" & input$category_gsea_atac == "C8"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_hs_go_atac <- msigdbr(species = "Homo sapiens", category = "C8")
print(v$msigdbr_hs_go_atac)
v$pathways_atac <- split(x = v$msigdbr_hs_go_atac$gene_symbol, f = v$msigdbr_hs_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
if(input$species_gsea_atac == "Mus musculus" & input$category_gsea_atac == "C8"){
v$atacData$predicted.id -> Idents(v$atacData)
v$msigdbr_mm_go_atac <- msigdbr(species = "Mus musculus", category = "C8")
print(v$msigdbr_mm_go_atac)
v$pathways_atac <- split(x = v$msigdbr_mm_go_atac$gene_symbol, f = v$msigdbr_mm_go_atac$gs_name)
print(v$pathways_atac)
v$markers_atac <- FindMarkers(v$atacData, ident.1 = input$gsea_atac.ct1, ident.2 = input$gsea_atac.ct2, min.pct = input$min_pct_atac1, logfc_atac.threshold = input$logfc_atac1, test.use_atac = input$test.use_atac1, assay = "RNA", min.cells.feature = 0, min.cells.group = 0)
v$markers_atac <- v$markers_atac %>% arrange(desc(avg_log2FC))
print(v$markers_atac)
v$markers_atac.log2FC <- v$markers_atac$avg_log2FC
names(v$markers_atac.log2FC) <- row.names(v$markers_atac)
v$markers_atac.log2FC <- sort(na.omit(v$markers_atac.log2FC), decreasing = TRUE)
print(v$markers_atac.log2FC)
v$fgsea_atacRes_atac <- fgsea(pathways = v$pathways_atac, stats = v$markers_atac.log2FC, eps = 0.0, minSize = 5, maxSize = 500) %>% arrange((padj))
print(v$fgsea_atacRes_atac)
v$topPathways_atacUp_atac <- v$fgsea_atacRes_atac[ES > 0][head(order(pval), n=10), pathway]
v$topPathways_atacDown_atac <- v$fgsea_atacRes_atac[ES < 0][head(order(pval), n=10), pathway]
v$topPathways_atac <- c(v$topPathways_atacUp_atac, rev(v$topPathways_atacDown_atac))
}
output$gsea_atac.done <- renderText(paste0("Gene set enrichment done!"))
v$isGSEAatacdone <- TRUE
})
}
})
output$gsea_atac.select <- renderUI({
if(is.null(v$pathways_atac)){
return(NULL)
}else{
selectInput("gsea_atac.pathway", label = "Gene set to visualise",
choices = names(v$pathways_atac))
}
})
output$gsea_atac_plot <- renderPlot({
if(is.null(v$pathways_atac) || is.null(v$isGSEAatacdone)){
return(NULL)
}else{
withProgress(message="Generating gsea_atac plot...", value=0, {
plotEnrichment(v$pathways_atac[[input$gsea_atac.pathway]], v$markers_atac.log2FC) + labs(title=input$gsea_atac.pathway)
})
}
})
output$gsea_atac_plot1 <- renderPlot({
if(is.null(v$pathways_atac) || is.null(v$isGSEAatacdone)){
return(NULL)
}else{
withProgress(message="Generating gsea_atac plot...", value=0, {
plotGseaTable(v$pathways_atac[v$topPathways_atac], v$markers_atac.log2FC, v$fgsea_atacRes_atac, gseaParam=0.2)
})
}
})
output$gsea_atac.table <- DT::renderDataTable(
v$fgsea_atacRes_atac, server = FALSE, options = list(scrollX = TRUE, scrollY = "400px"))
output$download_gsea_atac.table <- downloadHandler(
filename = function(){"gsea_atac Results.csv"},
content = function(fname){
withProgress(message="Downloading gsea_atac Results...", value=0, {
fwrite(v$fgsea_atacRes_atac, fname)
})
}
)
output$great_atac.ct1.select <- renderUI({
if(is.null(v$atac.markers)){
return(NULL)
}else{
selectInput("great_atac.ct1", label = "Celltype1",
choices = as.vector(v$atacData$predicted.id))
}
})
output$great_atac.ct2.select <- renderUI({
if(is.null(v$atac.markers)){
return(NULL)
}else{
selectInput("great_atac.ct2", label = "Celltype2",
choices = as.vector(v$atacData$predicted.id))
}
})
observeEvent(input$gsea_atac1, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Generating gene set enrichment analysis...", value=0, {
DefaultAssay(v$atacData) <- 'peaks'
v$atacData$predicted.id -> Idents(v$atacData)
v$da_peaks <- FindMarkers(object = v$atacData, ident.1 = input$great_atac.ct1, ident.2 = input$great_atac.ct2, logfc.threshold = input$logfc_great, min.pct = input$min_pct_great, test.use = input$test.use_atac2, min.cells.feature = 0, min.cells.group = 0)
v$da_peaks1 <- as.data.frame(rownames(v$da_peaks))
names(v$da_peaks1)[1] <- "position"
v$da_peaks1[c('chr', 'start', 'end')] <- str_split_fixed(v$da_peaks1$position, '-', 3)
v$da_peaks1[2:4]-> v$da_peaks2
makeGRangesFromDataFrame(v$da_peaks2) -> v$da_peaks3
granges(v$da_peaks3) -> v$da_peaks4
v$res_atac = great(v$da_peaks4, input$gene_set_atac, input$tss_atac)
print(v$res_atac)
v$tb = getEnrichmentTable(v$res_atac)
v$tb1 <- v$tb[order(v$tb$fold_enrichment, decreasing = T),][1:50,]
print(v$tb)
print(v$tb1)
output$gsea_atac1.done <- renderText(paste0("Gene set enrichment done!"))
v$isGSEAatac1done <- TRUE
})
}
})
output$great.select <- renderUI({
if(is.null(v$res_atac)){
return(NULL)
}else{
selectInput("great.pathway", label = "Pathway to visualise",
choices = v$tb$description)
}
})
output$great.plot <- renderPlot({
if(is.null(v$atacData) || is.null(v$isGSEAatac1done)){
plotly_empty()
}else{
withProgress(message="Generating Gene Set Enrichment Plot...", value=0, {
plotRegionGeneAssociations(v$res_atac, term_id = input$great.pathway)
})
}
})
output$great1.plot <- renderPlotly({
if(is.null(v$atacData) || is.null(v$isGSEAatac1done)){
plotly_empty()
}else{
withProgress(message="Generating Gene Set Enrichment Plot...", value=0, {
ggplot(v$tb1, aes(x = fold_enrichment, y = description)) + geom_point(aes(color=p_value_hyper,size=gene_set_size)) + scale_color_gradientn(colours = rainbow(5)) + labs(x='Fold Enrichment', y=NULL, color='P-value',size='Gene number') + theme( axis.title = element_text(face='bold'), axis.text = element_text(face='bold')) + theme_bw()
})
}
})
output$coverage.atac.select <- renderUI({
if(is.null(v$atac.markers)){
return(NULL)
}else{
selectInput("deg1.atac", label = "Peaks to visualise",
choices = rownames(v$atac.markers))
}
})
output$coverage.atac_feature.select <- renderUI({
if(is.null(v$atacData)){
return(NULL)
}else{
selectInput("atac_feature", label = "Genes to visualise",
choices = rownames(v$atacData@assays$RNA@counts), multiple = T)
}
})
output$coverage.plot <- renderPlot({
if(is.null(v$atac.markers)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
DefaultAssay(v$atacData) <- 'peaks'
CoveragePlot(
object = v$atacData,
region = input$deg1.atac,
features = input$atac_feature,
extend.upstream = 40000,
extend.downstream = 20000
)
})
}
})
observeEvent(input$doDeg_motif, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
pfm <- getMatrixSet(x = JASPAR2020, opts = list(collection = "CORE", tax_group = 'vertebrates', all_versions = FALSE))
v$atacData <- AddMotifs(object = v$atacData, genome = BSgenome.Hsapiens.UCSC.hg19, pfm = pfm)
v$da_peaks <- FindAllMarkers(object = v$atacData, only.pos = F, test.use = input$test.use_motif, min.pct = input$min_pct_motif, logfc.threshold = input$logfc_motif)
v$top.da.peak <- rownames(v$da_peaks[v$da_peaks$p_val < 0.005, ])
enriched.motifs <- FindMotifs(object = v$atacData, features = v$top.da.peak)
v$enriched.motifs <- enriched.motifs
shinyalert("DEG Analysis done", "DEG Analysis done, please perform data visualization", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$motif.select <- renderUI({
if(is.null(v$atacData)){
return(NULL)
}else{
selectInput("motif.atac", label = "Peaks to visualise",
choices = rownames(v$enriched.motifs))
}
})
output$motif.plot <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
MotifPlot(object = v$atacData, motifs = input$motif.atac)
})
}
})
observeEvent(input$calc_motif_activity, {
tpmFiles_atac <- input$tpmFiles_atac
tpmFiles_atac <- v$atacData
if (is.null(tpmFiles_atac)){
v$atacData <- NULL
shinyalert("Please upload your data", "Please upload your data", type = "warning", imageWidth = 10, imageHeight = 10)
}else{
withProgress(message="Finding DEGs...", value=0, {
v$atacData <- RunChromVAR(object = v$atacData, genome = BSgenome.Hsapiens.UCSC.hg19)
DefaultAssay(v$atacData) <- 'chromvar'
differential.activity <- FindAllMarkers(object = v$atacData, only.pos = F, mean.fxn = rowMeans, fc.name = "avg_diff")
v$differential.activity <- differential.activity
v$isMotifActivitydone <- TRUE
shinyalert("Motif analysis done", "Motif analysis done, please perform motif ", type = "success", imageWidth = 10, imageHeight = 10)
})
}
})
output$motif_feature.select <- renderUI({
if(is.null(v$atacData) || is.null(v$isMotifActivitydone)){
return(NULL)
}else{
selectInput("motif.feature", label = "Peaks to visualise",
choices = rownames(v$atacData))
}
})
output$motif_feature.plot <- renderPlotly({
if(is.null(v$atacData) || is.null(v$isMotifActivitydone)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
FeaturePlot(object = v$atacData,
features = input$motif.feature,
min.cutoff = 'q10',
max.cutoff = 'q90',
pt.size = 0.1)
})
}
})
output$motif1.select <- renderUI({
if(is.null(v$atacData)){
return(NULL)
}else{
selectInput("motif1.atac", label = "Peaks to visualise",
choices = rownames(v$differential.activity))
}
})
output$motif1.plot <- renderPlot({
if(is.null(v$atacData)){
return(NULL)
}else{
withProgress(message="Generating DEG Plot...", value=0, {
MotifPlot(object = v$atacData, motifs = input$motif1.atac, assay = 'peaks')
})
}
})
##---------------Summary tab
##------Clean up when ending session----
session$onSessionEnded(function(){
prePlot()
})
})
Raw
ui.R
This file has been truncated, but you can view the full file.
library(shiny)
library(plotly)
library(shinythemes)
library(shinydashboard)
library(shinyalert)
library(shinyjs)
library(shinyBS)
library(CellChat)
library(stringr)
library(fontawesome)
#library(hdf5r)
library(shinycssloaders)
library(Seurat)
library(pbmcapply)
library(devtools)
library(harmony)
library(reticulate)
library(sceasy)
library(pheatmap)
library(heatmaply)
library(liana)
library(rGREAT)
library(anndata)
library(clustree)
library(magrittr)
library(rhandsontable)
library(dplyr)
library(EnhancedVolcano)
library(ggplot2)
library(lisi)
library(cowplot)
library(FastIntegration)
library(Signac)
library(cluster)
library(GenomeInfoDb)
library(EnsDb.Hsapiens.v86)
library(BSgenome.Hsapiens.UCSC.hg19)
library(data.table)
library(purrr)
library(dplyr)
library(DT)
library(msigdbr)
library(fgsea)
library(MOFA2)
library(stats)
library(ktplots)
library(ggplot2)
library(SeuratWrappers)
library(shinyWidgets)
#library(car)
#library(nortest)
library(shinyFiles)
library(bslib)
#library(torch)
tags$style(type="text/css",
".shiny-output-error { visibility: hidden; }",
".shiny-output-error:before { visibility: hidden; }"
)
shinyUI(fluidPage(theme = shinytheme("cerulean"),
tagList(tags$head(tags$style(type = 'text/css','.navbar-brand{display:none;}')),
navbarPage("",
tabPanel(icon("home"),
titlePanel(h1(p(strong("ezSinglecell : An integrated one-stop single-cell and spatial analysis toolbox for bench scientists")), align = "center")),
fluidRow(
column(
br(),
p(strong("ezSingleCell"), "is an integrated one-stop single-cell and spatial analysis toolbox developed by Chen Jinmiao's lab with an intention to empower bench scientists to perform downstream Bioinformatics analysis. In the current version, we incorporate 5 modules : Single cell RNA-seq, Single cell Data Integration, Single cell Multiomics, Single Cell ATAC-seq and Spatial Transcriptomics.", style="text-align:justify;color:black;font-size:15px"),
column(6,
imageOutput("demo_image"),
),
column(6,
imageOutput("demo_image1"),
),
br(),
br(),
p("In this web server, we combine in-house novel algorithms such as CELLiD (for cell type identification), along with existing top performing methods for both basic and advanced downstream analyses such as batch effect removal, trajectory, cell-cell communication, differential abundance, and spatial deconvolution.", style="text-align:justify;color:black;font-size:15px"),
p("Currently ezSingleCell supports inputs in different formats such as text, csv or 10X cell ranger output. ezSingleCell is also available as a software package with a" , a("ShinyApp interface", href = "https://github.com/JinmiaoChenLab/ezSingleCell2", target = "_blank"), ",that can be run on a computer with basic memory requirements", style="text-align:justify;color:black;font-size:15px"),
p("The toolkit uses example data from", a("DISCO", href = "https://www.immunesinglecell.org/", target="_blank"),"that contains data from 13998 samples, covering 461 tissues/cell lines/organoids, 158 diseases, and 20 platforms.", style="text-align:justify;color:black;font-size:15px"), width=12)),
# p(em("Developed by"),br("CJM Lab"),style="text-align:center; font-family: times")
wellPanel(
HTML(
'<p align="center" width="4">Singapore Immunology Network, Agency for Science, Technology and Research (A*STAR)</p>
<p align="center" width="4">Github: <a href="https://github.com/JinmiaoChenLab/">https://github.com/JinmiaoChenLab/</a></p>
<p align="center" width="4">Created by <a href="https://www.a-star.edu.sg/sign/people/principal-investigators/jinmiao-chen">Jinmiao Chen Lab</a> and collaborated with <a href="https://www.vishuo.com/en/"> Vishuo Biomedical Pte. Ltd.</a></p>'
)),
),
tabPanel("Single cell RNA-Sequencing",
navlistPanel(widths=c(2,10),
tabPanel("Overview",
h2(p("Workflow for scRNA-Seq module")),
br(),
column(12,
imageOutput("scrna_image1"),
),
#column(6,
#imageOutput("scrna_image2")
#),
),
tabPanel("Upload your data",
column(9,
column(5,
#h4('Load Data:'),
wellPanel(
conditionalPanel(
condition = "input.scInput == 'Raw Counts Matrix'",
titlePanel(h4(p("Load your example data"))),
actionBttn("loadexample_tpm", "Load example and run", icon = icon("hand-o-right"), size = 'sm', onclick = "$(tab).removeClass('disabled')"),
#bsPopover("loadexample_tpm", "Load Example Data","Press to load example data (Raw counts matrix)", placement = "bottom", trigger = "hover", options = NULL)
),
conditionalPanel(
condition = "input.scInput == '10X cellranger'",
titlePanel(h4(p("Load your example data"))),
actionBttn("loadexample_scH5", "Load example and run", icon = icon("hand-o-right"), size = 'sm', onclick = "$(tab).removeClass('disabled')"),
#bsPopover("loadexample_scH5", "Load Example Data","Press to load example data (cellRanger output)", placement = "bottom", trigger = "hover", options = NULL)
),
conditionalPanel(
condition = "input.scInput == 'rds object'",
titlePanel(h4(p("Load your example data"))),
actionBttn("loadexample_rds", "Load example and run", icon = icon("hand-o-right"), size = 'sm', onclick = "$(tab).removeClass('disabled')"),
#bsPopover("loadexample_rds", "Load Example Data","Press to load example data (rds object)", placement = "bottom", trigger = "hover", options = NULL)
),
conditionalPanel(
condition = "input.scInput == 'h5ad'",
titlePanel(h4(p("Load your example data"))),
actionBttn("loadexample_h5ad", "Load example and run", icon = icon("hand-o-right"), size = 'sm', onclick = "$(tab).removeClass('disabled')"),
#bsPopover("loadexample_h5ad", "Load Example Data","Press to load example data (h5ad object)", placement = "bottom", trigger = "hover", options = NULL)
),
titlePanel(h4(HTML("<b>Load your input data</b>"))),
br(),
selectInput("scInput",
label = "Select Data Input Type",
choices = c("Raw Counts Matrix", "10X cellranger", "rds object", "h5ad"),
selected = "Raw Counts Matrix"),
#bsPopover("scInput", "Select Input Format","Users can select input format", placement = "bottom", trigger = "hover", options = NULL),
conditionalPanel(
condition = "input.scInput == 'Raw Counts Matrix'",
fileInput("tpmFiles",
label = "Counts File (Accepted Format: text)",
accept = ".txt"),
),
conditionalPanel(
condition = "input.scInput == '10X cellranger'",
fileInput("scH5",
label = "Cellranger output (Accepted Format: .h5)",
accept = ".h5"),
),
conditionalPanel(
condition = "input.scInput == 'rds object'",
fileInput("rds",
label = "Seurat Object (Accepted Format: .rds)",
accept = ".rds"),
),
conditionalPanel(
condition = "input.scInput == 'h5ad'",
fileInput("h5ad",
label = "AnnData Object (Accepted Format: .h5ad)",
accept = ".h5ad"),
),
#column(6,
# numericInput(inputId = "min.genes",
# label = "Min. genes",
# value = 200,
# min = 1)
# ),
#column(6,
# numericInput(inputId = "min.cells",
# label = "Min. cells",
# value = 3,
# min = 1)
# ),
textInput(inputId = "projName",
label = "Project Name",
value = "scRNA"),
fluidRow(
actionBttn("loadButton", "Load data", icon = icon("hand-o-right"), size = 'sm', onclick = "$(tab).removeClass('disabled')"),
#bsPopover("loadButton", "Load Data","Press to load your data", placement = "bottom", trigger = "hover", options = NULL),
actionBttn("reset_scRNA", "Reset", icon = icon("repeat"), size = 'sm'),
#bsPopover("reset_scRNA", "Reload Data","Press to reanalyze your data", placement = "bottom", trigger = "hover", options = NULL)
),
)),
chooseSliderSkin("Modern"),
titlePanel(h4(p("Quality control"))),
column(6,
plotOutput("nFeature_RNAPlot", width = "200%")
),
column(4,
br(),
downloadBttn('download_nFeature_RNA', 'Download (as png)', size = 'sm')
),
column(12,
column(3,
numericInput("ob1",
label = "Min nFeature:",
value = 200,
min = 0,
step = 1),
),
column(3,
numericInput("ob2",
label = "Max nFeature:",
value = 2500,
min = 0,
step = 1),
),
column(3,
numericInput("ob3",
label = "Mt%:",
value = 5,
min = 0,
step = 1),
),
column(3,
br(),
actionBttn("filter_seurat", "Filter", icon = icon("hand-o-right"), size = 'sm'),
#bsPopover("filter_seurat", "Filter Data","Press to filter data based on parameters", placement = "bottom", trigger = "hover", options = NULL),
),
),
),
column(12,
withSpinner(dataTableOutput('countdataDT')),
downloadBttn('downloadCount', 'Download Table'),
)),
#tabPanel("2. Quality control Plot",
# tabsetPanel(id="qc_scRNA",
# tabPanel("Violin Plot",
# column(3,
# plotOutput("nFeature_RNAPlot")
# ),
# column(3,
# plotOutput("mitoPlot")
# ),
# column(3,
# plotOutput("nCount_RNAPlot")
# ),
# column(12,
# column(3,
# downloadBttn('download_nFeature_RNA', 'Download nFeature (as png)', size = 'sm'),
# ),
# column(3,
# downloadBttn('download_mito', 'Download mito (as png)', size = 'sm'),
# ),
# column(3,
# downloadBttn('download_nCount_RNA', 'Download nCount (as png)', size = 'sm'),
# ),
# )
# ),
# tabPanel("Feature Scatter Plot",
# column(5,
# plotlyOutput("FeatureScatterPlot1")
# ),
# column(5,
# plotlyOutput("FeatureScatterPlot2")
# ),
# br(),
# br(),
# column(12,
# column(5,
# downloadBttn('download_FeatureScatterPlot1', 'Download (as png)', size = 'sm'),
# ),
# column(5,
# downloadBttn('download_FeatureScatterPlot2', 'Download (as png)', size = 'sm'),
# ),
# ),
# )
# ),
# ),
tabPanel("Normalization and Variable Feature Selection", value = "test",
tags$script(
'
var tab = $(\'a[data-value="test"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
selectInput("norm1",
label = "Normalization method",
choices = c("LogNormalize", "SCTransform"),
selected = "SCTransform"
),
textOutput("nVarGenes"),
fluidRow(
column(3,
numericInput("var.genes",
label = "Number of variable genes",
value = 2000,
min = 500,
step = 500)
),
column(3,
selectInput("selection.method",
label = "Selection method",
choices = c("vst", "dispersion"))
),
column(4,
br(),
actionBttn("findVarGenes", "Identify highly variable genes", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab1).removeClass('disabled')"),
#bsPopover("findVarGenes", "Feature Selection","Press to identify highly variable features", placement = "bottom", trigger = "hover", options = NULL),
#actionButton("doSCTransform", "Run SCTransform", icon = icon("hand-pointer-o"))
# actionButton("doVarplot", "Plot variable genes", icon = icon("hand-pointer-o"))
)),
plotOutput("VarGenes", width = "100%"),
),
tabPanel( "PCA", value = "test1",
tags$script(
'
var tab1 = $(\'a[data-value="test1"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
tabsetPanel(id="Pca",
tabPanel(title="PCA Plot", value="P_panel1",
br(),
#column(3,
#selectInput("assays1",
# label = "Normalization method:",
# choices = c("LogNormalization", "SCTransform"),
# selected = "SCTransform"
# ),
# ),
br(),
column(3,
actionBttn("doPCA", "Run PCA", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab2).removeClass('disabled')"),
#bsPopover("doPCA", "Run PCA","Press to run PCA Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
br(),
br(),
column(6,
plotlyOutput("PCA2DPlot", width = "100%")
),
column(12,
column(3,
br(),
downloadBttn('download_PCA', 'Download PCA Plot (as png)', size = 'sm'),
),
column(3,
br(),
downloadBttn('download_PCA_embedding', 'Download PCA Embedding (as csv)', size = 'sm'),
),
),
),
tabPanel(title="PC Gene Visualisation", value="P_panel2",
br(),
selectInput("select.pc",
label = "PC to plot",
choices = c(1:50)
),
fluidRow(
column(4,
plotOutput("vizPlot", width = "100%", height = "600px")
),
column(8,
plotOutput("PCHeatmap", width = "100%", height = "600px")
),
column(12,
column(3,
br(),
downloadBttn('download_vizPlot', 'Download vizPlot (as png)', size = 'sm'),
),
column(3,
br(),
downloadBttn('download_PCHeatmap', 'Download PCHeatmap (as png)', size = 'sm'),
),
#br(),
DT::dataTableOutput("PCtable"),
column(3,
downloadBttn('download_PCTable', 'Download top genes (as csv)', size = 'sm'),
),
),
),
),
tabPanel(title="Elbow", value="P_panel4",
br(),
#actionButton("doElbow", label = "Get Elbow Plot"),
#br(),
br(),
plotOutput("Elbow", width = "100%"),
column(12,
column(3,
downloadBttn('download_Elbow', 'Download ElbowPlot (as png)', size = 'sm'),
),
),
)
)),
tabPanel( "UMAP", value = "test2",
tags$script(
'
var tab2 = $(\'a[data-value="test2"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
#titlePanel(h4(p("UMAP Analysis"))),
br(),
fluidRow(
column(3,
numericInput("dim.used",
label = "Dimensions used",
value = 10)
),
br(),
column(3,
actionBttn("doUmap", "Run UMAP", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab3).removeClass('disabled')"),
#bsPopover("doUmap", "Run UMAP","Press to run UMAP Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("Umap.done"),
br()
)),
br(),
plotlyOutput("Umap_2d_plot_1", width = "50%"),
br(),
column(12,
column(3,
downloadBttn('download_UMAP', 'Download UMAP Plot (as png)', size = 'sm'),
),
column(3,
downloadBttn('download_UMAP_embedding', 'Download UMAP Embeddings (as csv)', size = 'sm'),
),
),
),
tabPanel( "tSNE", value = "test2",
tags$script(
'
var tab2 = $(\'a[data-value="test2"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
numericInput("dim.used",
label = "Dimensions used",
value = 10)
),
column(3,
uiOutput("perplex.option")
),
column(3,
br(),
actionBttn("doTsne", "Run TSNE", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab3).removeClass('disabled')"),
#bsPopover("doTsne", "Run tSNE Analysis","Press to run tSNE Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("Tsne.done"),
br()
)),
br(),
plotlyOutput("Tsne_2d_plot_1", width = "50%"),
br(),
column(12,
column(3,
downloadBttn('download_Tsne', 'Download tSNE Plot (as png)', size = 'sm'),
),
column(3,
downloadBttn('download_Tsne_embedding', 'Download tSNE Embeddings (as csv)', size = 'sm'),
),
),
),
tabPanel("Clustering", value = "test2",
tags$script(
'
var tab2 = $(\'a[data-value="test2"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
tabsetPanel(id="cluster",
tabPanel(title="Clustering", value="C_panel1",
br(),
fluidRow(
column(12,
column(2,
numericInput("clus.res",
label = "Resolution used",
value = 0.6,
min = 0.1,
step = 0.1)
),
column(2,
selectInput("dim.used",
label = "PC to use",
choices = c(10:50)),
),
column(2,
br(),
actionBttn("findCluster", "Find Clusters", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab3).removeClass('disabled')"),
#bsPopover("findCluster", "Run Clustering","Press to run Cluster Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("cluster.done"),
br()
),
),
column(6,
plotlyOutput("Cluster2DPlot_1", width = "100%"),
br(),
),
),
column(6,
column(6,
downloadBttn('download_Cluster', 'Download ClusterPlot (as png)', size = 'sm'),
),
column(6,
downloadBttn('download_ClusterTable', 'Download Cluster table (as csv)', size = 'sm'),
),
),
),
tabPanel(title="Determine cluster resolution", value="C_panel2",
br(),
column(9,
h4(p("Determine cluster resolution:")),
column(4,
numericInput("clus.res_a",
label = "Resolution (from)",
value = 0.6,
min = 0.1,
step = 0.1)
),
column(4,
numericInput("clus.res_b",
label = "Resolution (to)",
value = 1,
min = 0.1,
step = 0.1)
),
column(4,
br(),
actionBttn("findoptimumCluster", "Determine optimum resolution", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("findoptimumCluster", "Determine optimum resolution","Press to determine optimum cluster resolution", placement = "bottom", trigger = "hover", options = NULL),
),
br(),
column(12,
br(),
br(),
br(),
br(),
plotOutput("OptimumCluster2DPlot_1", width = "100%"),
br(),
),
column(12,
column(6,
downloadBttn('download_OptimumCluster', 'Download ClusterPlot (as png)', size = 'sm'),
),
column(6,
downloadBttn('download_OptimumClusterTable', 'Download Cluster table (as csv)', size = 'sm'),
),
),
),
),
tabPanel(title="Subcluster Analysis", value="C_panel3",
column(12,
titlePanel(h4(p("Subcluster Analysis"))),
),
column(3,
br(),
uiOutput("subcluster.gene.select"),
),
column(3,
br(),
numericInput("subcluster.res",
label = "Resolution used",
value = 0.6,
min = 0.1,
step = 0.1),
#bsTooltip("subcluster", "Perform Subcluster Analysis", placement = "bottom", trigger = "hover",
#options = NULL)
),
column(3,
br(),
br(),
actionBttn("subcluster", "Run Subcluster analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("subcluster", "Subcluster Analysis","Press to perform Subcluster Analysis for a specific cluster at any given resolution", placement = "bottom", trigger = "hover", options = NULL),
textOutput("Subcluster.done"),
br()
),
br(),
br(),
br(),
column(9,
plotlyOutput("subcluster_plot")
),
column(12,
column(4,
downloadBttn('download_subcluster', 'Download Subcluster plot (as png)', size = 'sm'),
br(),
),
),
),
),
),
tabPanel("Cell type Identification", value = "test3",
tags$script(
'
var tab3 = $(\'a[data-value="test3"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
selectInput("cellid_method",
label = "Celltype annotation method",
choices = c("CELLiD", "Celltypist"),
selected = "CELLiD"),
),
),
tabsetPanel(id="ct1",
tabPanel(title="Celltype Identification", value="Ct_panel1",
conditionalPanel(
condition = "input.cellid_method == 'CELLiD'",
column(3,
selectInput("cellatlas",
label = "Reference Atlas",
choices = c("all", "adipose", "adrenal_gland", "blood", "bone_marrow", "brain", "breast", "breast_milk", "eye", "gut", "heart", "kidney", "liver", "lung", "pancreas", "PDAC", "skin", "testis", "thymus", "tonsil"),
selected = "all")
),
column(3,
br(),
actionBttn("doCELLiD", "Run CELLiD", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab4).removeClass('disabled')"),
#bsPopover("doCELLiD", "Perform Celltype Identification","Press to run celltype identification using CELLiD", placement = "bottom", trigger = "hover", options = NULL),
textOutput("CELLiD.done"),
br()
),
br(),
br(),
br(),
column(12,
column(6,
plotlyOutput("Umap_cellid", width = "100%"),
),
column(6,
plotlyOutput("Umap_cellid1", width = "100%"),
),
),
DT::dataTableOutput("ct.table"),
column(12,
column(4,
downloadBttn('download_Umap_cellid', 'Download CELLiD predict1 (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_Umap_cellid1', 'Download CELLiD predict2 (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_cellid_prediction', 'Download CELLiD predictions (in csv)', size = 'sm'),
br(),
),
),
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler3",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 2){
console.log("got here");
$("a:contains(Spatial Transcriptomics)").click();
}
});
')),
column(3,
br(),
actionBttn("dodeconv_spatial", "Go to Spatial Deconvolution", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("dodeconv_spatial", "Perform Celltype Deconvolution","Press to navigate to Spatial module and perform celltype deconvolution", placement = "bottom", trigger = "hover", options = NULL),
),
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler5",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 2){
console.log("got here");
$("a:contains(Single cell ATAC-seq)").click();
}
});
')),
column(3,
br(),
actionBttn("doct_atac", "Annotate cell types for ATAC-data", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("doct_atac", "Perform Celltype Label Transfer","Press to navigate to scATAC-seq module and perform celltype label transfer", placement = "bottom", trigger = "hover", options = NULL),
),
),
conditionalPanel(
condition = "input.cellid_method == 'Celltypist'",
column(3,
selectInput("celltypistatlas",
label = "Reference Atlas",
choices = c("Immune_All_Low.pkl", "Autopsy_COVID19_Lung.pkl", "Pan_Fetal_Human.pkl", "Nuclei_Lung_Airway.pkl", "Developing_Human_Thymus.pkl", "Human_Lung_Atlas.pkl", "Developing_Mouse_Brain.pkl", "Developing_Human_Brain.pkl", "Cells_Lung_Airway.pkl", "Healthy_COVID19_PBMC.pkl", "Human_IPF_Lung.pkl", "Adult_Mouse_Gut.pkl", "Immune_All_High.pkl", "COVID19_Immune_Landscape.pkl", "Human_PF_Lung.pkl", "COVID19_HumanChallenge_Blood.pkl", "Lethal_COVID19_Lung.pkl", "Cells_Fetal_Lung.pkl", "Cells_Intestinal_Tract.pkl"),
selected = "Immune_All_Low.pkl")
),
column(3,
br(),
actionBttn("doCelltypist", "Run Celltypist", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab4).removeClass('disabled')"),
#bsPopover("doCelltypist", "Perform Celltype Identification","Press to Press to run celltype identification using CellTypist", placement = "bottom", trigger = "hover", options = NULL),
textOutput("Celltypist.done"),
br()
),
br(),
br(),
br(),
column(12,
column(6,
plotlyOutput("Umap_celltypist", width = "100%"),
),
column(6,
plotlyOutput("Umap_celltypist1", width = "100%"),
),
),
DT::dataTableOutput("celltypist.table"),
#DT::dataTableOutput(outputId = "recoding"),
#DT::dataTableOutput(outputId = "newVars")
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler3a",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 2){
console.log("got here");
$("a:contains(Spatial Transcriptomics)").click();
}
});
')),
column(3,
br(),
actionBttn("dodeconv_spatial1", "Go to Spatial Deconvolution", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("dodeconv_spatial1", "Perform Celltype Deconvolution","Press to navigate to Spatial module and perform celltype deconvolution", placement = "bottom", trigger = "hover", options = NULL),
),
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler5a",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 2){
console.log("got here");
$("a:contains(Single cell ATAC-seq)").click();
}
});
')),
column(3,
br(),
actionBttn("doct_atac1", "Annotate cell types for ATAC-data", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("doct_atac1", "Perform Celltype Label Transfer","Press to navigate to scATAC-seq module and perform celltype label transfer", placement = "bottom", trigger = "hover", options = NULL),
),
),
),
tabPanel(title="Rename Clusters", value="Ct_panel2",
conditionalPanel(
condition = "input.cellid_method == 'CELLiD'",
column(12,
column(6,
br(),
h4(p("Please modify your celltypes below:")),
br(),
actionBttn("commitButton", "Rename clusters (if needed)", size = "sm"),
#bsPopover("commitButton", "Rename clusters","Press to rename clusters", placement = "bottom", trigger = "hover", options = NULL),
br(),
br(),
rHandsontableOutput("hot"),
),
column(6,
h4(p("UMAP with renamed celltypes")),
plotlyOutput("Umap_cellid2", width = "100%"),
br(),
),
),
),
conditionalPanel(
condition = "input.cellid_method == 'Celltypist'",
br(),
h3(p("Not supported yet"))
# column(12,
# column(6,
# br(),
# h4(p("Please modify your celltypes below:")),
# br(),
# #actionBttn("commitButton1", "Rename clusters", size = "sm"),
# br(),
# br(),
#rHandsontableOutput("cot"),
# ),
# column(6,
# h4(p("UMAP with renamed celltypes")),
# plotOutput("Umap_celltypist2", width = "100%"),
# br(),
# ),
#),
# ),
),
),
tabPanel(title="Visualize", value="Ct_panel3",
conditionalPanel(
condition = "input.cellid_method == 'CELLiD'",
column(12,
column(6,
br(),
uiOutput("ct.gene.select"),
actionBttn("Vis_seurat1", "Visualize", icon = icon("hand-pointer-o"), size = 'sm'),
plotlyOutput("ct.gene1.plot", width = "100%"),
br(),
),
column(6,
br(),
br(),
plotlyOutput("ct.gene.plot", width = "150%"),
br(),
),
),
column(12,
column(4,
downloadBttn('download_violn1', 'Download Violin plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_feature1', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
),
),
conditionalPanel(
condition = "input.cellid_method == 'Celltypist'",
column(6,
br(),
uiOutput("celltypist.gene.select"),
plotlyOutput("celltypist.gene1.plot", width = "100%"),
br(),
),
column(6,
plotlyOutput("celltypist.gene.plot", width = "150%"),
br(),
),
),
),
),
),
tabPanel("Cell type Similarity", value = "test4",
tags$script(
'
var tab4 = $(\'a[data-value="test4"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(3,
br(),
selectInput("cell1",
label = "Group by",
choices = c("seurat_clusters", "primary.predict", "newID"),
selected = "primary.predict"),
),
column(3,
br(),
selectInput("corr_method",
label = "Statistics",
choices = c("pearson", "spearman", "kendall"),
selected = "pearson"),
),
column(3,
br(),
br(),
actionBttn("cell_cell", "Run celltype similarity", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab5).removeClass('disabled')"),
#bsPopover("cell_cell", "Perform Celltype Similarity","Press to run Celltype Similarity Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("CELL.done"),
br()
),
br(),
br(),
br(),
column(9,
plotlyOutput("cell_cell_sim")
),
br(),
br(),
br(),
column(12,
column(4,
downloadBttn('download_cell_cell_sim', 'Download Celltype similarity plot (as png)', size = 'sm'),
),
column(4,
downloadBttn('download_cor.table', 'Download Celltype similarity table (in csv)', size = 'sm'),
),
),
br(),
br(),
br(),
br(),
br(),
DT::dataTableOutput("cor.table")
),
tabPanel("DEGs", value = "test5",
tags$script(
'
var tab5 = $(\'a[data-value="test5"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
selectInput("deg_method",
label = "Type of DEG analysis",
choices = c("Celltype specific", "Pairwise DEGs"),
selected = "Celltype specific"),
),
),
conditionalPanel(
condition = "input.deg_method == 'Celltype specific'",
fluidRow(
column(3,
selectInput("deg1",
label = "Group by",
choices = c("seurat_clusters", "primary.predict", "newID"),
selected = "primary.predict"),
),
column(3, numericInput("min_pct",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, numericInput("logfc",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, selectInput("test.use",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
),
br(),
column(3,
actionBttn("doDeg", "Run DEGs", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab6).removeClass('disabled')"),
#bsPopover("doDeg", "Perform DEG Analysis","Press to run DEG Analysis", placement = "bottom", trigger = "hover", options = NULL),
)),
br(),
column(12,
DT::dataTableOutput("Deg.table"),
br(),
br(),
br(),
plotOutput("Deg3.plot", width = "100%")
)
),
conditionalPanel(
condition = "input.deg_method == 'Pairwise DEGs'",
#br(),
column(12,
column(3,
selectInput("deg3",
label = "Group by",
choices = c("seurat_clusters", "primary.predict", "newID"),
selected = "primary.predict"),
),
column(3,
uiOutput("gene1.select"),
),
column(3,
uiOutput("gene2.select"),
),
br(),
column(3,
actionBttn("doVolcano", "Run Pairwise DEGs", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab6).removeClass('disabled')"),
#bsPopover("doVolcano", "Perform Pairwise DEG Analysis","Press to run Pairwise DEG Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
),
column(12,
column(3, numericInput("min_pct_a",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, numericInput("logfc_a",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, selectInput("test.use_a",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
)
),
column(12,
plotOutput("volcano.plot", width = "75%"),
),
br(),
column(12,
plotOutput("dega.plot", width = "100%")
),
),
),
tabPanel("Data visualization", value = "test5",
tags$script(
'
var tab5 = $(\'a[data-value="test5"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
selectInput("deg2",
label = "Group by",
choices = c("seurat_clusters", "primary.predict", "newID"),
selected = "primary.predict"),
),
column(6,
uiOutput("deg.gene.select"),
actionBttn("Vis_seurat", "Visualize", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab6).removeClass('disabled')"),
plotlyOutput("Deg.plot", width = "150%"),
br(),
plotlyOutput("Deg1.plot", width = "150%"),
br(),
plotOutput("Deg2.plot", width = "150%"),
),
column(4,
downloadBttn('download_violn', 'Download Violin plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_feature', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_ridge', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
)
),
tabPanel("GSEA", value = "test6",
tags$script(
'
var tab6 = $(\'a[data-value="test6"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(12,
br(),
column(2, selectInput("species_gsea",
label = "Species",
choices = c("Homo sapiens", "Mus musculus"))
),
column(2, selectInput("category_gsea",
label = "Collection",
choices = c("H", "C2", "C5", "C7", "C8"))
),
column(2,
uiOutput("gsea.ct1.select"),
),
column(2,
uiOutput("gsea.ct2.select"),
),
column(2, numericInput("min_pct1",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(2, numericInput("logfc1",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
),
column(12,
column(2, selectInput("test.use1",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
),
column(4,
uiOutput("gsea.select"),
),
br(),
column(3,
actionBttn("gsea", "Run gene set enrichment analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("gsea", "Perform Gene Set Enrichment Analysis","Press to run Gene Set Enrichment Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
textOutput("gsea.done"),
br()
),
br(),
column(7,
plotOutput("gsea_plot", width = "100%"),
),
br(),
br(),
column(12,
plotOutput("gsea_plot1", width = "100%"),
),
br(),
br(),
DT::dataTableOutput("gsea.table"),
column(4,
downloadBttn('download_gsea.table', 'Download GSEA Results (in csv)', size = 'sm'),
),
),
tabPanel("Cell-cell communication", value = "test6",
tags$script(
'
var tab6 = $(\'a[data-value="test6"]\').parent().addClass("disabled");
$(function(){
$(tab.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
br(),
selectInput("cc1a",
label = "Group by",
choices = c("seurat_clusters", "primary.predict"),
selected = "primary.predict"),
),
column(3,
br(),
selectInput("cc_method",
label = "Method",
choices = c("natmi", "connectome", "logfc", "sca", "cellphonedb", "cytotalk"), selected = "cellphonedb")
),
column(3,
br(),
selectInput("cc_resource",
label = "Method",
choices = c("Default", "Consensus", "Baccin2019", "CellCall", "CellChatDB", "Cellinker", "CellPhoneDB", "CellTalkDB", "connectomeDB2020", "EMBRACE", "Guide2Pharma", "HPMR", "ICELLNET", "iTALK", "Kirouac2010", "LRdb", "Ramilowski2015", "OmniPath"), selected = "CellPhoneDB")
),
column(3,
br(),
br(),
actionBttn("doCC", "Run Analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("doCC", "Perform Cell-Cell Communication Analysis","Press to run Cell-Cell Communication Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("cc.done"),
br()
),
),
br(),
br(),
column(12,
plotOutput("CC_plot1", width = "100%"),
),
column(3,
uiOutput("CC.gene.select"),
),
column(3,
uiOutput("CC.gene1.select"),
),
br(),
column(12,
plotOutput("CC_plot2", width = "100%"),
),
column(12,
h4(p("Interacting partners")),
),
withSpinner(dataTableOutput('cc.table')),
br(),
#column(12,
# h4(p("p-values for the all the interacting partners")),
# ),
# withSpinner(dataTableOutput('cc.table1')),
),
#tabPanel("13. Annotate scATAC-seq data using scRNA-seq data",
# tags$head(tags$script('
# Shiny.addCustomMessageHandler("myCallbackHandler1",
# function(typeMessage) {console.log(typeMessage)
# if(typeMessage == 2){
# console.log("got here");
# $("a:contains(Single cell ATAC-seq)").click();
# }
# });
# ')),
#
# column(5,
# br(),
# actionBttn("loadexample_atacseq", "Load example scATAC data and annotate", icon = icon("hand-pointer-o"), size = 'sm'),
# #bsPopover("loadexample_atacseq", "Load example scATAC data and annotate","Press to load example scATAC data and annotate", placement = "bottom", trigger = "hover", options = NULL),
# ),
# column(4,
# br(),
# actionBttn("process_atacseq", "Process scATAC-seq data", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("process_atacseq", "Process scATAC-seq data","Press to navigate to scATAC-seq module and process", placement = "bottom", trigger = "hover", options = NULL),
# ),
# br(),
# column(12,
# plotOutput("annotate_scRNA_ATAC_plot", width = "70%"),
# plotOutput("annotate_scRNA_ATAC_plot1", width = "70%")
# ),
# )
),
),
##------------Data Integration module--------------------##
tabPanel("Single cell data integration",
navlistPanel(widths=c(2,10),
tabPanel("Overview",
h2(p("Workflow for Data Integration module")),
br(),
imageOutput("integration_image"),
),
tabPanel("Upload your data",
column(9,
column(5,
#h4('Load Data:'),
wellPanel(
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
titlePanel(h4(p("Load your example data"))),
actionBttn("loadexample1", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_intg).removeClass('disabled')"),
),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
titlePanel(h4(p("Load your example data"))),
actionBttn("loadexample1a", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_intg).removeClass('disabled')"),
),
titlePanel(h4(p("Load your input data"))),
br(),
selectInput("scInput1",
label = "Select Data Input Type",
choices = c("Raw Counts Matrix", "10X cellranger"),
selected = "Raw Counts Matrix"),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
fileInput("tpmFiles1",
label = "Counts File (Accepted Format: text)",
accept = ".txt",
multiple = T),
fileInput("cellAnnoFiles1",
label = "Upload Metadata (Accepted Format: tab delimited text)",
accept = ".txt",
multiple = T),
),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
fileInput("scH5_1",
label = "Cellranger output (Accepted Format: .h5)",
accept = ".h5",
multiple = T),
),
column(6,
numericInput(inputId = "min.genes1",
label = "Min. genes",
value = 200,
min = 1)
),
column(6,
numericInput(inputId = "min.cells1",
label = "Min. cells",
value = 3,
min = 1)
),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix' || input.scInput1 == '10X cellranger'",
textInput(inputId = "projName1",
label = "Project Name",
value = "Integration")),
fluidRow(
actionBttn("loadButton1", "Load data", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_intg).removeClass('disabled')"),
actionBttn("reset_intg", "Reset", icon = icon("repeat"), size = "sm")
),
)),
chooseSliderSkin("Modern"),
titlePanel(h4(p("Quality control"))),
column(6,
plotOutput("nFeature_RNAPlot1", width = "200%")
),
column(12,
column(3,
numericInput("ob1a",
label = "Min nFeature:",
value = 200,
min = 0,
step = 1),
),
column(3,
numericInput("ob2a",
label = "Max nFeature:",
value = 2500,
min = 0,
step = 1),
),
column(3,
numericInput("ob3a",
label = "Mt%:",
value = 5,
min = 0,
step = 1),
),
column(3,
br(),
actionBttn("filter_seurat1", "Filter", icon = icon("hand-o-right"), size = "sm")
),
),
),
column(12,
withSpinner(dataTableOutput('countdataDT1'))
)),
#tabPanel("2. Quality control",
# tabsetPanel(id="qc_intg",
# tabPanel("Violin Plot",
# column(3,
# plotOutput("nFeature_RNAPlot1")
# ),
# column(3,
# plotOutput("mitoPlot1")
# ),
# column(3,
# plotOutput("nCount_RNAPlot1")
# )),
# tabPanel("Feature Scatter",
# column(5,
# plotlyOutput("FeatureScatterPlot1a")
# ),
# column(5,
# plotlyOutput("FeatureScatterPlot2a")
# ),
# )),
# ),
tabPanel("Normalization and Variable Feature Selection", value = "test_intg",
tags$script(
'
var tab_intg = $(\'a[data-value="test_intg"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
textOutput("nVarGenes_bef_intg"),
fluidRow(
column(3,
numericInput("var.genes_bef_intg",
label = "Number of variable genes",
value = 2000,
min = 500,
step = 500)
),
column(3,
selectInput("selection.method_bef_intg",
label = "Selection method",
choices = c("vst", "dispersion"))
),
column(4,
br(),
actionBttn("findVarGenes_bef_intg", "Identify highly variable genes", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg1).removeClass('disabled')"),
)),
plotOutput("VarGenes_bef_intg", width = "100%")
),
tabPanel("Before Data Integration", value = "test_intg1",
tags$script(
'
var tab_intg1 = $(\'a[data-value="test_intg1"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
tabsetPanel(id="bef_data_integration",
tabPanel("PCA",
br(),
tabsetPanel(id="Pca_bef_intg_seurat",
tabPanel(title="PCA Plot",
br(),
fluidRow(
column(3,
actionBttn("runPCA_bef_intg", "Run PCA", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg2).removeClass('disabled')")
),
),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
plotlyOutput("PCAplot_bef_tpm1", width = "75%"),
plotlyOutput("PCAplot_bef_tpm2", width = "75%"),
plotlyOutput("PCAplot_bef_tpm3", width = "75%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
plotlyOutput("PCAplot_bef_h5_1", width = "75%"),
plotlyOutput("PCAplot_bef_h5_2", width = "75%")),
),
tabPanel(title="PC Gene Visualisation",
br(),
selectInput("select.pc_bef_intg",
label = "PC to plot",
choices = c(1:50)
),
fluidRow(
column(4,
plotOutput("vizPlot_bef_intg", width = "100%", height = "600px")
),
column(8,
plotOutput("PCHeatmap_bef_intg", width = "100%", height = "600px")
)
),
DT::dataTableOutput("PCtable_bef_intg")
),
tabPanel(title="Elbow",
br(),
br(),
plotOutput("Elbow_bef_intg", width = "100%")
)),
),
tabPanel("UMAP",
br(),
column(3,
numericInput("dim.used_bef_intg",
label = "Dimensions used",
value = 10)
),
column(9,
br(),
actionBttn("runUMAP_bef_intg", "Run UMAP", icon = icon("hand-pointer-o"), size = "sm"),
br(),
br(),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
plotlyOutput("UMAPplot_bef_tpm1", width = "75%"),
br(),
plotlyOutput("UMAPplot_bef_tpm2", width = "75%"),
br(),
plotlyOutput("UMAPplot_bef_tpm3", width = "75%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
plotlyOutput("UMAPplot_bef_h5_1", width = "75%"),
br(),
plotlyOutput("UMAPplot_bef_h5_2", width = "75%")),
),
),
tabPanel("tSNE",
br(),
column(3,
numericInput("dim.used_bef_intg",
label = "Dimensions used",
value = 10)
),
column(9,
br(),
actionBttn("runTSNE_bef_intg", "Run TSNE", icon = icon("hand-pointer-o"), size = "sm"),
#textOutput("Intg.done"),
br(),
br(),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
plotlyOutput("TSNEplot_bef_tpm1", width = "75%"),
br(),
plotlyOutput("TSNEplot_bef_tpm2", width = "75%"),
br(),
plotlyOutput("TSNEplot_bef_tpm3", width = "75%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
plotlyOutput("TSNEplot_bef_h5_1", width = "75%"),
br(),
plotlyOutput("TSNEplot_bef_h5_2", width = "75%")),
),
),
#tabPanel("Clustering",
# br(),
# fluidRow(
# column(3,
# numericInput("clus.res_bef_intg",
# label = "Resolution used",
# value = 0.6,
# min = 0.1,
# step = 0.1)
# ),
# column(3,
# selectInput("dim.used_bef_intg",
# label = "PC to use",
# choices = c(10:50)),
# ),
# column(3,
# br(),
# actionBttn("findCluster_bef_intg", "Find Clusters", icon = icon("hand-pointer-o"), size = "sm"),
# #textOutput("cluster1.done"),
# ),
# br(),
# br(),
# br(),
# br(),
# plotlyOutput("Cluster2DPlot_bef_intg", width = "75%")
# ),
# column(12,
# h4(p("Determine cluster resolution:")),
# column(3,
# numericInput("clus.res_a1",
# label = "Resolution (from)",
# value = 0.6,
# min = 0.1,
# step = 0.1)
# ),
# column(3,
# numericInput("clus.res_b1",
# label = "Resolution (to)",
# value = 1,
# min = 0.1,
# step = 0.1)
# ),
# column(4,
# br(),
# actionBttn("findoptimumCluster1", "Determine optimum resolution", icon = icon("hand-pointer-o"), size = 'sm'),
# #bsPopover("findoptimumCluster", "Determine optimum resolution","Press to determine optimum cluster resolution", placement = "bottom", trigger = "hover", options = NULL),
# ),
# ),
# br(),
# column(9,
# br(),
# plotOutput("OptimumCluster2DPlot_a1", width = "100%"),
# br(),
# ),
# column(6,
# column(6,
# downloadBttn('download_OptimumCluster1', 'Download ClusterPlot (as png)', size = 'sm'),
# ),
# column(6,
# downloadBttn('download_OptimumClusterTable1', 'Download Cluster table (as csv)', size = 'sm'),
# ),
# ),
# column(12,
# titlePanel(h4(p("Subcluster Analysis"))),
# ),
# column(3,
# br(),
# uiOutput("subcluster.gene.select1"),
# ),
# column(3,
# br(),
# numericInput("subcluster.res1",
# label = "Resolution used",
# value = 0.6,
# min = 0.1,
# step = 0.1),
# #bsTooltip("subcluster", "Perform Subcluster Analysis", placement = "bottom", trigger = "hover",
# #options = NULL)
# ),
# column(3,
# br(),
# br(),
# actionBttn("subcluster1", "Run Subcluster analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("subcluster", "Subcluster Analysis","Press to perform Subcluster Analysis for a specific cluster at any given resolution", placement = "bottom", trigger = "hover", options = NULL),
# textOutput("Subcluster1.done"),
# br()
# ),
# br(),
# br(),
# br(),
# column(9,
# plotlyOutput("subcluster_plot1")
# ),
# column(12,
# column(4,
# downloadBttn('download_subcluster1', 'Download Subcluster plot (as png)', size = 'sm'),
# br(),
# ),
# ),
#),
#tabPanel("Cell type identification",
#conditionalPanel(
# condition = "input.scAnalysis_integ == 'Seurat'",
# br(),
# column(3,
# selectInput("cellid_method1",
# label = "Celltype annotation method",
# choices = c("CELLiD", "Celltypist"),
# selected = "CELLiD"),
# ),
# conditionalPanel(
# condition = "input.cellid_method1 == 'CELLiD'",
# column(3,
# selectInput("cellatlas1",
# label = "Reference Atlas",
# choices = c("all", "adipose", "adrenal_gland", "blood", "bone_marrow", "brain", "breast", "breast_milk", "eye", "gut", "heart", "kidney", "liver", "lung", "pancreas", "PDAC", "skin", "testis", "thymus", "tonsil"),
# selected = "all")
# ),
# column(3,
# br(),
# actionBttn("doCELLiD_bef_intg", "Run CELLiD", icon = icon("hand-pointer-o"), size = "sm"),
# textOutput("CELLiD1.done"),
# br()
# ),
# br(),
# br(),
# br(),
# br(),
# br(),
# plotlyOutput("Umap_cellid_bef_intg", width = "50%"),
# plotlyOutput("Umap_cellid_bef_intg1", width = "50%"),
# br(),
# br(),
# DT::dataTableOutput("ct_bef_intg.table")
#),
# ),
# conditionalPanel(
# condition = "input.cellid_method1 == 'Celltypist'",
# column(3,
# selectInput("celltypistatlas1",
# label = "Reference Atlas",
# choices = c("Immune_All_Low.pkl", "Autopsy_COVID19_Lung.pkl", "Pan_Fetal_Human.pkl", "Nuclei_Lung_Airway.pkl", "Developing_Human_Thymus.pkl", "Human_Lung_Atlas.pkl", "Developing_Mouse_Brain.pkl", "Developing_Human_Brain.pkl", "Cells_Lung_Airway.pkl", "Healthy_COVID19_PBMC.pkl", "Human_IPF_Lung.pkl", "Adult_Mouse_Gut.pkl", "Immune_All_High.pkl", "COVID19_Immune_Landscape.pkl", "Human_PF_Lung.pkl", "COVID19_HumanChallenge_Blood.pkl", "Lethal_COVID19_Lung.pkl", "Cells_Fetal_Lung.pkl", "Cells_Intestinal_Tract.pkl"),
# selected = "Immune_All_Low.pkl")
# ),
# column(3,
# br(),
# actionBttn("doCelltypist_bef_intg", "Run Celltypist", icon = icon("hand-pointer-o"), size = 'sm'),
# textOutput("Celltypist1.done"),
# br()
# ),
# br(),
# br(),
# br(),
# br(),
# br(),
# plotlyOutput("Umap_celltypist_bef_intg", width = "50%"),
# plotlyOutput("Umap_celltypist_bef_intg1", width = "50%"),
# br(),
# br(),
# DT::dataTableOutput("ct_celltypist_bef_intg.table")
# ),
# )
)
),
tabPanel("Data Integration", value = "test_intg2",
tags$script(
'
var tab_intg2 = $(\'a[data-value="test_intg2"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(12,
br(),
column(3,
selectInput("scAnalysis_integ",
label = "Analysis method",
choices = c("Seurat", "Harmony", "scVI", "fastMNN"),
selected = "Seurat"),
),
#br(),
conditionalPanel(
condition = "input.scAnalysis_integ == 'Seurat'",
column(4,
numericInput("nfeatures_intg_seurat",
label = "Integration features",
value = 2000,
min = 500,
step = 1)),
column(4,
br(),
actionBttn("doIntg_seurat", "Run Data Integration", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg3).removeClass('disabled')"),
),
),
conditionalPanel(
condition = "input.scAnalysis_integ == 'Harmony'",
column(4,
numericInput("nfeatures_intg_harmony",
label = "Integration features",
value = 2000,
min = 500,
step = 1)),
br(),
column(4,
actionBttn("doIntg_harmony", "Run Data Integration", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg3).removeClass('disabled')"),
),
),
conditionalPanel(
condition = "input.scAnalysis_integ == 'scVI'",
column(3,
numericInput("nfeatures_intg_scvi",
label = "Integration features",
value = 2000,
min = 500,
step = 1),
),
column(3,
numericInput("scvi_latent",
label = "Number of dimensions",
value = 10,
min = 1,
step = 1),
),
column(3,
br(),
actionBttn("doIntg_scvi", "Run Data Integration", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg3).removeClass('disabled')"),
),
column(12,
h4('This may take long time to run (around 30 mins to 1 hr)'),
),
br(),
),
conditionalPanel(
condition = "input.scAnalysis_integ == 'fastMNN'",
column(4,
numericInput("nfeatures_intg_fastmnn",
label = "Integration features",
value = 2000,
min = 500,
step = 1)),
br(),
column(4,
actionBttn("doIntg_fastmnn", "Run Data Integration", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg3).removeClass('disabled')"),
)),
))),
tabPanel("PCA", value = "test_intg3",
tags$script(
'
var tab_intg3 = $(\'a[data-value="test_intg3"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_integ == 'Seurat'",
tabsetPanel(id="Pca1",
tabPanel(title="PCA Plot",
br(),
column(3,
actionBttn("runPCA_intg_seurat", "Run PCA", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg4).removeClass('disabled')"),
),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
br(),
br(),
br(),
plotlyOutput("PCAplot_seurat_tpm1", width = "75%"),
plotlyOutput("PCAplot_seurat_tpm2", width = "75%"),
plotlyOutput("PCAplot_seurat_tpm3", width = "75%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
br(),
br(),
br(),
plotlyOutput("PCAplot_seurat_h5_1", width = "75%"),
plotlyOutput("PCAplot_seurat_h5_2", width = "75%")),
),
tabPanel(title="PC Gene Visualisation",
br(),
selectInput("select.pc_intg_seurat",
label = "PC to plot",
choices = c(1:50)
),
fluidRow(
column(4,
plotOutput("vizPlot_intg_seurat", width = "100%", height = "600px")
),
column(8,
plotOutput("PCHeatmap_intg_seurat", width = "100%", height = "600px")
)
),
DT::dataTableOutput("PCtable_intg_seurat")
),
tabPanel(title="Elbow", value="P_panel4",
br(),
br(),
plotOutput("Elbow_intg_seurat", width = "100%")
))),
conditionalPanel(
condition = "input.scAnalysis_integ == 'Harmony'",
tabsetPanel(id="Pca1",
tabPanel(title="PCA Plot",
br(),
column(3,
actionBttn("runPCA_intg_harmony", "Run PCA", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg4).removeClass('disabled')"),
),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
br(),
br(),
br(),
plotlyOutput("PCAplot_harmony_tpm1", width = "75%"),
plotlyOutput("PCAplot_harmony_tpm2", width = "75%"),
plotlyOutput("PCAplot_harmony_tpm3", width = "75%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
br(),
br(),
br(),
plotlyOutput("PCAplot_harmony_h5_1", width = "75%"),
plotlyOutput("PCAplot_harmony_h5_2", width = "75%")),
),
tabPanel(title="PC Gene Visualisation",
br(),
selectInput("select.pc_intg_harmony",
label = "PC to plot",
choices = c(1:50)
),
fluidRow(
column(4,
plotOutput("vizPlot_intg_harmony", width = "100%", height = "600px")
),
column(8,
plotOutput("PCHeatmap_intg_harmony", width = "100%", height = "600px")
)
),
DT::dataTableOutput("PCtable_intg_harmony")
),
tabPanel(title="Elbow", value="P_panel4",
br(),
br(),
plotOutput("Elbow_intg_harmony", width = "100%")
))),
conditionalPanel(
condition = "input.scAnalysis_integ == 'scVI'",
tabsetPanel(id="Pca2",
tabPanel(title="PCA Plot",
br(),
column(3,
actionBttn("runPCA_intg_scvi", "Run PCA", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg4).removeClass('disabled')"),
),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
br(),
br(),
br(),
plotlyOutput("PCAplot_scvi_tpm1", width = "75%"),
plotlyOutput("PCAplot_scvi_tpm2", width = "75%"),
plotlyOutput("PCAplot_scvi_tpm3", width = "75%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
br(),
br(),
br(),
plotlyOutput("PCAplot_scvi_h5_1", width = "75%"),
plotlyOutput("PCAplot_scvi_h5_2", width = "75%")),
),
)),
conditionalPanel(
condition = "input.scAnalysis_integ == 'fastMNN'",
tabsetPanel(id="Pca2",
tabPanel(title="PCA Plot",
br(),
column(3,
actionBttn("runPCA_intg_fastmnn", "Run PCA", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg4).removeClass('disabled')"),
),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
br(),
br(),
br(),
plotlyOutput("PCAplot_fastmnn_tpm1", width = "100%"),
plotlyOutput("PCAplot_fastmnn_tpm2", width = "100%"),
plotlyOutput("PCAplot_fastmnn_tpm3", width = "100%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
br(),
br(),
br(),
plotlyOutput("PCAplot_fastmnn_h5_1", width = "100%"),
plotlyOutput("PCAplot_fastmnn_h5_2", width = "100%")),
),
tabPanel(title="PC Gene Visualisation",
br(),
selectInput("select.pc_intg_fastmnn",
label = "PC to plot",
choices = c(1:50)
),
fluidRow(
column(4,
plotOutput("vizPlot_intg_fastmnn", width = "100%", height = "600px")
),
column(8,
DT::dataTableOutput("PCtable_intg_fastmnn")
)),
),
))
),
tabPanel("UMAP", value = "test_intg4",
tags$script(
'
var tab_intg4 = $(\'a[data-value="test_intg4"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(3,
numericInput("dim.used_intg",
label = "Dimensions used",
value = 10, step = 1)
),
column(9,
br(),
actionBttn("runUMAP_intg", "Run UMAP", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg5).removeClass('disabled')"),
br(),
br(),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
plotlyOutput("UMAPplot_intg_tpm1", width = "100%"),
br(),
plotlyOutput("UMAPplot_intg_tpm2", width = "100%"),
br(),
plotlyOutput("UMAPplot_intg_tpm3", width = "100%"),
br(),
textOutput("UMAP_lisi_intg")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
plotOutput("UMAPplot_intg_h5_1", width = "100%"),
br(),
plotOutput("UMAPplot_intg_h5_2", width = "100%")),
),
),
tabPanel("tSNE", value = "test_intg4",
tags$script(
'
var tab_intg4 = $(\'a[data-value="test_intg4"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(3,
numericInput("dim.used_intg1",
label = "Dimensions used",
value = 10)
),
column(9,
br(),
actionBttn("runTSNE_intg", "Run TSNE", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg5).removeClass('disabled')"),
#textOutput("Intg.done"),
br(),
br(),
conditionalPanel(
condition = "input.scInput1 == 'Raw Counts Matrix'",
plotlyOutput("TSNEplot_intg_tpm1", width = "100%"),
br(),
plotlyOutput("TSNEplot_intg_tpm2", width = "100%"),
br(),
plotlyOutput("TSNEplot_intg_tpm3", width = "100%")),
br(),
conditionalPanel(
condition = "input.scInput1 == '10X cellranger'",
plotlyOutput("TSNEplot_intg_h5_1", width = "100%"),
br(),
plotlyOutput("TSNEplot_intg_h5_2", width = "100%")),
),
),
tabPanel("Clustering", value = "test_intg4",
tags$script(
'
var tab_intg4 = $(\'a[data-value="test_intg4"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
numericInput("clus.res_intg",
label = "Resolution used",
value = 0.6,
min = 0.1,
step = 0.1)
),
column(3,
selectInput("dim.used_intg2",
label = "PC to use", selected = "30",
choices = c(1:50)),
),
column(3,
br(),
actionBttn("findCluster_intg", "Find Clusters", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg5).removeClass('disabled')"),
textOutput("cluster1.done"),
),
br(),
br(),
br(),
br(),
plotlyOutput("Cluster2DPlot_intg", width = "50%")
),
column(12,
h4(p("Determine cluster resolution:")),
column(3,
numericInput("clus.res_a2",
label = "Resolution (from)",
value = 0.6,
min = 0.1,
step = 0.1)
),
column(3,
numericInput("clus.res_b2",
label = "Resolution (to)",
value = 1,
min = 0.1,
step = 0.1)
),
column(4,
br(),
actionBttn("findoptimumCluster2", "Determine optimum resolution", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("findoptimumCluster", "Determine optimum resolution","Press to determine optimum cluster resolution", placement = "bottom", trigger = "hover", options = NULL),
),
column(12,
column(9,
plotOutput("OptimumCluster2DPlot_b1", width = "100%"),
br(),
),
),
column(12,
column(6,
downloadBttn('download_OptimumCluster2', 'Download ClusterPlot (as png)', size = 'sm'),
),
column(6,
downloadBttn('download_OptimumClusterTable2', 'Download Cluster table (as csv)', size = 'sm'),
),
),
column(12,
titlePanel(h4(p("Subcluster Analysis"))),
),
column(3,
br(),
uiOutput("subcluster.gene.select2"),
),
column(3,
br(),
numericInput("subcluster.res2",
label = "Resolution used",
value = 0.6,
min = 0.1,
step = 0.1),
#bsTooltip("subcluster", "Perform Subcluster Analysis", placement = "bottom", trigger = "hover",
#options = NULL)
),
column(3,
br(),
br(),
actionBttn("subcluster2", "Run Subcluster analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("subcluster", "Subcluster Analysis","Press to perform Subcluster Analysis for a specific cluster at any given resolution", placement = "bottom", trigger = "hover", options = NULL),
textOutput("Subcluster2.done"),
br()
),
br(),
br(),
br(),
column(9,
plotlyOutput("subcluster_plot2")
),
column(12,
column(4,
downloadBttn('download_subcluster2', 'Download Subcluster plot (as png)', size = 'sm'),
br(),
),
),
),
),
tabPanel("Cell type identification",value = "test_intg5",
tags$script(
'
var tab_intg5 = $(\'a[data-value="test_intg5"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
selectInput("cellid_method5",
label = "Celltype annotation method",
choices = c("CELLiD", "Celltypist"),
selected = "CELLiD"),
)
),
conditionalPanel(
condition = "input.cellid_method5 == 'CELLiD'",
column(3,
#br(),
selectInput("cellatlas1a",
label = "Reference Atlas",
choices = c("all", "adipose", "adrenal_gland", "blood", "bone_marrow", "brain", "breast", "breast_milk", "eye", "gut", "heart", "kidney", "liver", "lung", "pancreas", "PDAC", "skin", "testis", "thymus", "tonsil"),
selected = "all")
),
column(3,
br(),
actionBttn("doCELLiD_intg", "Run CELLiD", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg6).removeClass('disabled')"),
textOutput("CELLiD2.done"),
br()
),
br(),
br(),
br(),
plotlyOutput("Umap_cellid_intg", width = "50%"),
plotlyOutput("Umap_cellid_intg1", width = "50%"),
br(),
br(),
DT::dataTableOutput("ct_intg.table")
),
conditionalPanel(
condition = "input.cellid_method5 == 'Celltypist'",
column(3,
#br(),
selectInput("celltypistatlas5",
label = "Reference Atlas",
choices = c("Immune_All_Low.pkl", "Autopsy_COVID19_Lung.pkl", "Pan_Fetal_Human.pkl", "Nuclei_Lung_Airway.pkl", "Developing_Human_Thymus.pkl", "Human_Lung_Atlas.pkl", "Developing_Mouse_Brain.pkl", "Developing_Human_Brain.pkl", "Cells_Lung_Airway.pkl", "Healthy_COVID19_PBMC.pkl", "Human_IPF_Lung.pkl", "Adult_Mouse_Gut.pkl", "Immune_All_High.pkl", "COVID19_Immune_Landscape.pkl", "Human_PF_Lung.pkl", "COVID19_HumanChallenge_Blood.pkl", "Lethal_COVID19_Lung.pkl", "Cells_Fetal_Lung.pkl", "Cells_Intestinal_Tract.pkl"),
selected = "Immune_All_Low.pkl")
),
column(3,
br(),
actionBttn("doCelltypist_intg", "Run Celltypist", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg6).removeClass('disabled')"),
textOutput("Celltypist5.done"),
br()
),
br(),
br(),
br(),
#br(),
#br(),
plotlyOutput("Umap_celltypist_intg", width = "50%"),
plotlyOutput("Umap_celltypist_intg1", width = "50%"),
br(),
br(),
DT::dataTableOutput("ct_celltypist_intg.table")
),
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler6",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 2){
console.log("got here");
$("a:contains(Spatial Transcriptomics)").click();
}
});
')),
column(3,
br(),
actionBttn("dodeconv_spatial_intg", "Go to Spatial Deconvolution", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("dodeconv_spatial", "Perform Celltype Deconvolution","Press to navigate to Spatial module and perform celltype deconvolution", placement = "bottom", trigger = "hover", options = NULL),
),
#br(),
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler7",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 2){
console.log("got here");
$("a:contains(Single cell ATAC-seq)").click();
}
});
')),
column(3,
br(),
actionBttn("doct_atac_intg", "Annotate cell types for ATAC-data", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("doct_atac", "Perform Celltype Label Transfer","Press to navigate to scATAC-seq module and perform celltype label transfer", placement = "bottom", trigger = "hover", options = NULL),
),
),
tabPanel("Cell type similarity", value = "test_intg6",
tags$script(
'
var tab_intg6 = $(\'a[data-value="test_intg6"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(3,
br(),
selectInput("cell1a",
label = "Group by",
choices = c("seurat_clusters", "primary.predict", "celltype"),
selected = "primary.predict"),
),
column(3,
br(),
selectInput("corr_method1",
label = "Statistics",
choices = c("pearson", "spearman", "kendall"),
selected = "pearson"),
),
column(3,
br(),
br(),
actionBttn("cell_cell1", "Run celltype similarity", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg7).removeClass('disabled')"),
#bsPopover("cell_cell", "Perform Celltype Similarity","Press to run Celltype Similarity Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("CELL1.done"),
br()
),
br(),
br(),
br(),
column(9,
plotlyOutput("cell_cell_sim1")
),
br(),
br(),
br(),
column(12,
column(4,
downloadBttn('download_cell_cell_sim1', 'Download Celltype similarity plot (as png)', size = 'sm'),
),
column(4,
downloadBttn('download_cor.table1', 'Download Celltype similarity table (in csv)', size = 'sm'),
),
),
br(),
br(),
br(),
br(),
br(),
DT::dataTableOutput("cor.table1")
),
tabPanel("DEGs", value = "test_intg7",
tags$script(
'
var tab_intg7 = $(\'a[data-value="test_intg7"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
selectInput("deg_method1",
label = "Type of DEG analysis",
choices = c("Celltype specific", "Pairwise DEGs"),
selected = "Celltype specific"),
),
),
conditionalPanel(
condition = "input.deg_method1 == 'Celltype specific'",
fluidRow(
column(3,
selectInput("deg1a",
label = "Group by",
choices = c("seurat_clusters", "Predicted celltype", "Metadata celltype"),
selected = "Predicted celltype"),
),
column(3, numericInput("min_pct_intg",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, numericInput("logfc_intg",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, selectInput("test.use_intg",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
),
column(4,
actionBttn("doDeg_intg", "Run DEGs", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_intg8).removeClass('disabled')")
)),
br(),
column(12,
DT::dataTableOutput("Deg.table_intg")
)),
conditionalPanel(
condition = "input.deg_method1 == 'Pairwise DEGs'",
#br(),
column(12,
column(3,
selectInput("deg3a",
label = "Group by",
choices = c("seurat_clusters", "Predicted celltype", "Metadata celltype"),
selected = "primary.predict"),
),
column(3,
uiOutput("gene1a.select"),
),
column(3,
uiOutput("gene2a.select"),
),
br(),
column(3,
actionBttn("doVolcano1", "Run Pairwise DEGs", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg8).removeClass('disabled')"),
#bsPopover("doVolcano", "Perform Pairwise DEG Analysis","Press to run Pairwise DEG Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
),
column(12,
column(3, numericInput("min_pct_a1",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, numericInput("logfc_a1",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, selectInput("test.use_a1",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
)
),
column(12,
plotOutput("volcano.plot1", width = "75%"),
),
br(),
column(12,
plotOutput("dega1.plot", width = "100%")
),
),
),
tabPanel("Data Visualization", value = "test_intg7",
tags$script(
'
var tab_intg7 = $(\'a[data-value="test_intg7"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(3,
selectInput("deg1b",
label = "Group by",
choices = c("seurat_clusters", "Predicted celltype", "Metadata celltype"),
selected = "Predicted celltype"),
),
column(6,
uiOutput("deg.gene.select_intg"),
actionBttn("Vis_intg", "Visualize", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_intg8).removeClass('disabled')"),
plotlyOutput("Deg.plot_intg", width = "200%"),
br(),
plotlyOutput("Deg1.plot_intg", width = "100%"),
br(),
plotOutput("Deg2.plot_intg", width = "200%")
),
column(4,
downloadBttn('download_violn_intg', 'Download Violin plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_feature_intg', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_ridge_intg', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
),
tabPanel("GSEA", value = "test_intg8",
tags$script(
'
var tab_intg8 = $(\'a[data-value="test_intg8"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(12,
br(),
column(2, selectInput("species_gsea1",
label = "Species",
choices = c("Homo sapiens", "Mus musculus"))
),
column(2, selectInput("category_gsea1",
label = "Collection",
choices = c("H", "C2", "C5", "C7", "C8"))
),
column(2,
uiOutput("gsea.ct1a.select"),
),
column(2,
uiOutput("gsea.ct2a.select"),
),
column(2, numericInput("min_pct1a",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(2, numericInput("logfc1a",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
),
column(12,
column(2, selectInput("test.use1a",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
),
column(4,
uiOutput("gsea1.select"),
),
br(),
column(3,
actionBttn("gsea1", "Run gene set enrichment analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("gsea", "Perform Gene Set Enrichment Analysis","Press to run Gene Set Enrichment Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
textOutput("gsea1.done"),
br()
),
br(),
column(7,
plotOutput("gsea_plot1a", width = "100%"),
),
br(),
br(),
column(12,
plotOutput("gsea_plot1b", width = "100%"),
),
br(),
br(),
DT::dataTableOutput("gsea.table1"),
column(4,
downloadBttn('download_gsea.table1', 'Download GSEA Results (in csv)', size = 'sm'),
),
),
tabPanel("Cell-cell communication", value = "test_intg8",
tags$script(
'
var tab_intg8 = $(\'a[data-value="test_intg8"]\').parent().addClass("disabled");
$(function(){
$(tab_intg.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
fluidRow(
column(3,
br(),
selectInput("cc1b",
label = "Group by",
choices = c("seurat_clusters", "primary.predict"),
selected = "primary.predict"),
),
column(3,
br(),
selectInput("cc_method1",
label = "Method",
choices = c("natmi", "connectome", "logfc", "sca", "cellphonedb", "cytotalk"), selected = "cellphonedb")
),
column(3,
br(),
selectInput("cc_resource1",
label = "Method",
choices = c("Default", "Consensus", "Baccin2019", "CellCall", "CellChatDB", "Cellinker", "CellPhoneDB", "CellTalkDB", "connectomeDB2020", "EMBRACE", "Guide2Pharma", "HPMR", "ICELLNET", "iTALK", "Kirouac2010", "LRdb", "Ramilowski2015", "OmniPath"), selected = "CellPhoneDB")
),
column(3,
br(),
br(),
actionBttn("doCC1", "Run analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("doCC", "Perform Cell-Cell Communication Analysis","Press to run Cell-Cell Communication Analysis", placement = "bottom", trigger = "hover", options = NULL),
textOutput("cc1.done"),
br()
),
),
column(12,
plotOutput("CC_plot1a", width = "100%"),
),
column(3,
uiOutput("CC.gene.select1"),
),
column(3,
uiOutput("CC.gene1.select1"),
),
br(),
br(),
column(12,
plotOutput("CC_plot2a", width = "100%"),
),
column(12,
h4(p("Interacting partners")),
),
withSpinner(dataTableOutput('cc.table1a')),
),
#tabPanel("13. Annotate scATAC-seq data using scRNA-seq data",
# tags$head(tags$script('
# Shiny.addCustomMessageHandler("myCallbackHandler1a",
# function(typeMessage) {console.log(typeMessage)
# if(typeMessage == 2){
# console.log("got here");
# $("a:contains(Single cell ATAC-seq)").click();
# }
# });
# ')),
# column(5,
# br(),
# actionBttn("loadexample_atacseq1", "Load example scATAC data and annotate", icon = icon("hand-pointer-o"), size = 'sm'),
# #bsPopover("loadexample_atacseq", "Load example scATAC data and annotate","Press to load example scATAC data and annotate", placement = "bottom", trigger = "hover", options = NULL),
# ),
# column(4,
# br(),
# actionBttn("process_atacseq1", "Process scATAC-seq data", icon = icon("hand-pointer-o"), size = 'sm'),
# #bsPopover("process_atacseq", "Process scATAC-seq data","Press to navigate to scATAC-seq module and process", placement = "bottom", trigger = "hover", options = NULL),
# ),
# br(),
# column(12,
# plotOutput("annotate_scRNA_ATAC_plot_a", width = "70%"),
# plotOutput("annotate_scRNA_ATAC_plot_1a", width = "70%")
# ),
#)
),
),
##------------Spatial Transcriptomics module--------------------##
tabPanel("Spatial Transcriptomics",
navlistPanel(widths=c(2,10),
tabPanel("Overview",
h2(p("Workflow for Spatial Transcriptomics module")),
br(),
imageOutput("spatial_image"),
),
tabPanel("Upload your data",
column(9,
column(5,
#h4('Load Data:'),
wellPanel(
conditionalPanel(
titlePanel(h4(p("Load your example data"))),
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'Seurat'",
actionBttn("loadexample_seurat_spatial", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_sp).removeClass('disabled')"),
),
conditionalPanel(
titlePanel(h4(p("Load your example data"))),
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'GraphST'",
actionBttn("loadexample_graphst_spatial", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_sp).removeClass('disabled')"),
),
conditionalPanel(
titlePanel(h4(p("Load your example data"))),
condition = "input.scAnalysis_platform == 'Xenium' & input.scAnalysis_sp == 'Seurat'",
actionBttn("loadexample_xenium", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_sp).removeClass('disabled')"),
),
titlePanel(h4(HTML("<b>Load your input data</b>"))),
br(),
selectInput("scInput3",
label = "Select Data Input Type",
choices = c("SpaceRanger output"),
selected = "SpaceRanger output"),
selectInput("scAnalysis_platform",
label = "Platform",
choices = c("Visium", "Xenium"),
selected = "Visium"),
selectInput("scAnalysis_sp",
label = "Analysis method",
choices = c("Seurat", "GraphST"),
selected = "Seurat"),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'Seurat'",
fileInput("tpmFile_spatial",
label = "Upload H5 output from Spaceranger (Accepted Format: .h5)",
accept = ".h5"),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'GraphST'",
fileInput("tpmFile_graphst",
label = "Upload H5 output from Spaceranger (Accepted Format: .h5)",
accept = ".h5"),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'Seurat'",
shinyFiles::shinyDirButton(id = 'dir', label = "Path to spaceRanger output file", title = "Sheets Folder Selector"),
verbatimTextOutput("dir", placeholder = TRUE),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'GraphST'",
shinyFiles::shinyDirButton(id = 'dir_graphst', label = "Path to spaceRanger output file", title = "Sheets Folder Selector"),
verbatimTextOutput("dir_graphst", placeholder = TRUE),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
shinyFiles::shinyDirButton(id = 'dir_xenium', label = "Path to xenium files", title = "Sheets Folder Selector"),
verbatimTextOutput("dir_xenium", placeholder = TRUE),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'Seurat'",
textInput(inputId = "projName3",
label = "Project Name",
value = "Spatial_seurat"),
fluidRow(
actionBttn("loadButton3", "Load Data", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_sp).removeClass('disabled')"),
actionBttn("reset3", "Reset Data", icon = icon("repeat"), size = "sm"),
)),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'GraphST'",
textInput(inputId = "projName3a",
label = "Project Name",
value = "Spatial_graphST"),
fluidRow(
actionBttn("loadButton3a", "Load Data", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_sp).removeClass('disabled')"),
actionBttn("reset3a", "Reset Data", icon = icon("repeat"), size = "sm"),
)),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
#actionButton("loadexample_xenium", "Load example and run", icon = icon("hand-o-right")),
textInput(inputId = "projName_xenium",
label = "Project Name",
value = "Xenium"),
fluidRow(
actionBttn("load_xenium", "Load Data", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp).removeClass('disabled')"),
actionBttn("reset_xenium", "Reset Data", icon = icon("repeat"), size = "sm"),
)),
)),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium'",
chooseSliderSkin("Modern"),
titlePanel(h4(p("Quality control"))),
column(7,
plotOutput("nCount_SpatialPlot", width = "200%"),
),
column(7,
plotOutput("SpatialFeaturePlot", width = "200%"),
column(4,
numericInput("obs",
label = "nMolecules:",
value = 200,
min = 0,
step = 1),
),
column(4,
numericInput("obs1",
label = "nGenes:",
value = 200,
min = 0,
step = 1),
),
column(3,
numericInput("obs2",
label = "Mt%:",
value = 30,
min = 0,
step = 1),
),
column(6,
actionBttn("filter_spatial", "Filter", icon = icon("hand-pointer-o"), size = "sm"),
#actionButton("findVarGenes_mult", "Identify highly variable genes", icon = icon("hand-pointer-o")),
)
#column(12,
# sliderInput("obs", "Number of molecules:", min = 0, max = 30000, value = 200),
#),
#column(12,
# sliderInput("obs1", "Number of genes:", min = 0, max = 30000, value = 200),
#),
#column(12,
# sliderInput("obs2", "Mitochondrial percentage:", min = 0, max = 100, value = 30),
#),
),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
column(3,
br(),
numericInput("count_xenium",
label = "nCount_Xenium",
value = 0,
min = 0,
step = 1)
),
),
),
column(12,
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'Seurat'",
h4(p("Spatial counts")),
withSpinner(dataTableOutput('countdataDT_spatial')),
h4(p("H&E Image")),
withSpinner(plotOutput("h_e", width = "100%")),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
h4(p("Spatial counts")),
withSpinner(dataTableOutput('countdataDT_xenium')),
h4(p("H&E Image")),
withSpinner(plotOutput("h_e_xenium", width = "100%")),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium' & input.scAnalysis_sp == 'GraphST'",
h4(p("Spatial counts")),
withSpinner(dataTableOutput('countdataDT_spatial_graphst')),
h4(p("H&E Image")),
withSpinner(plotOutput("h_e_graphst", width = "100%")),
),
),
),
#tabPanel("2. Quality Control",
#conditionalPanel(
# condition = "input.scAnalysis_platform == 'Xenium'",
#column(12,
#column(5,
# plotOutput("nFeature_xenium")
#),
# column(5,
# plotOutput("nCount_xenium")
#),
#)
#),
#),
#tabPanel("3. FeatureScatter Plot",
# conditionalPanel(
# condition = "input.scAnalysis_platform == 'Visium'",
# column(6,
# plotlyOutput("FeatureScatterPlot_sp")
# ),
# ),
# conditionalPanel(
# condition = "input.scAnalysis_platform == 'Xenium'",
# column(6,
# plotlyOutput("FeatureScatterPlot_xenium")
# ),
# ),
#),
tabPanel("Normalization and Variable Feature Selection", value = "test_sp",
tags$script(
'
var tab_sp = $(\'a[data-value="test_sp"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium'",
fluidRow(
column(3,
numericInput("var.genes_sp",
label = "Number of variable genes",
value = 2000,
min = 500,
step = 500)
),
column(4,
br(),
actionBttn("doSCTransform_sp", "Run scTransform", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp1).removeClass('disabled')"),
#actionButton("findVarGenes_mult", "Identify highly variable genes", icon = icon("hand-pointer-o")),
)),
plotOutput("VarGenes_sp", width = "100%"),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
fluidRow(
column(3,
numericInput("var.genes_xenium",
label = "Number of variable genes",
value = 2000,
min = 500,
step = 500)
),
column(4,
br(),
actionBttn("doSCTransform_xenium", "Run scTransform", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp1).removeClass('disabled')"),
#actionButton("findVarGenes_mult", "Identify highly variable genes", icon = icon("hand-pointer-o")),
)),
plotOutput("VarGenes_xenium", width = "100%"),
),
),
#tabPanel("Gene expression visualization", value = "test_sp1",
# tags$script(
# '
# var tab_sp1 = $(\'a[data-value="test_sp1"]\').parent().addClass("disabled");
# $(function(){
# $(tab_sp.parent()).on("click", "li.disabled", function(e) {
# e.preventDefault();
# return false;
# });
# });
# '
# ),
# conditionalPanel(
# condition = "input.scAnalysis_platform == 'Visium'",
# fluidRow(
# column(6,
# uiOutput("sp.gene.select"),
# plotOutput("sp.plot", width = "100%"),
# ),
# )),
# conditionalPanel(
# condition = "input.scAnalysis_platform == 'Xenium'",
# column(6,
# uiOutput("xenium.gene.select"),
# plotOutput("markerplot_xenium")
# ),
# column(6,
# uiOutput("xenium.gene1.select"),
# plotOutput("featureplot_xenium")
# ),
# br(),
# br(),
# column(2,
# numericInput("x1_xenium",
# label = "x1",
# value = 1200,
# min = 0,
# step = 1),
# ),
# column(2,
# numericInput("x2_xenium",
# label = "x2",
# value = 2800,
# min = 0,
# step = 1),
# ),
# column(2,
# numericInput("y1_xenium",
# label = "y1",
# value = 3700,
# min = 0,
# step = 1),
# ),
# column(2,
# numericInput("y2_xenium",
# label = "y2",
# value = 4500,
# min = 0,
# step = 1),
# ),
# column(2,
# br(),
# actionBttn("crop_xenium", "Crop", icon = icon("hand-o-right")),
# ),
# fluidRow(
# column(6,
# uiOutput("xenium.gene2.select"),
# plotOutput("zoom_xenium")
# ),
# ),
# ),
#),
tabPanel("PCA", value = "test_sp1",
tags$script(
'
var tab_sp1 = $(\'a[data-value="test_sp1"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium'",
tabsetPanel(id="Pca_sp",
tabPanel(title="PCA Plot",
br(),
fluidRow(
column(3,
actionBttn("runPCA_spatial", "Run PCA", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp2).removeClass('disabled')")
),
),
br(),
plotlyOutput("PCAplot_spatial", width = "50%"),
),
tabPanel(title="PC Gene Visualisation",
br(),
selectInput("select.pc_spatial",
label = "PC to plot",
choices = c(1:50)
),
fluidRow(
column(4,
plotOutput("vizPlot_spatial", width = "100%", height = "600px")
),
column(8,
plotOutput("PCHeatmap_spatial", width = "100%", height = "600px")
)
),
DT::dataTableOutput("PCtable_spatial")
),
tabPanel(title="Elbow",
br(),
br(),
plotOutput("Elbow_spatial", width = "100%")
))),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
tabsetPanel(id="Pca_xenium",
tabPanel(title="PCA Plot",
br(),
fluidRow(
column(3,
actionBttn("runPCA_xenium", "Run PCA", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp2).removeClass('disabled')")
),
),
br(),
plotlyOutput("PCAplot_xenium", width = "100%"),
),
tabPanel(title="Elbow",
br(),
br(),
plotOutput("Elbow_xenium", width = "100%")
))
),
),
tabPanel("UMAP", value = "test_sp2",
tags$script(
'
var tab_sp2 = $(\'a[data-value="test_sp2"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium'",
column(3,
numericInput("dim.used_spatial",
label = "Dimensions used",
value = 10)
),
column(3,
br(),
actionBttn("runUMAP_spatial", "Run UMAP", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp3).removeClass('disabled')")
),
),
conditionalPanel(
condition = "input.scAnalysis_sp1 == 'Seurat' & input.scAnalysis_platform == 'Visium'",
column(12,
br(),
br(),
plotlyOutput("DimPlot_spatial", width = "75%")
)),
conditionalPanel(
condition = "input.scAnalysis_sp1 == 'GraphST' & input.scAnalysis_platform == 'Visium'",
column(12,
br(),
br(),
plotlyOutput("DimPlot_GraphST", width = "75%")
),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
column(3,
selectInput("dim.used_xenium",
label = "Dimensions",
choices = c(10:50),
selected = 30),
),
column(3,
br(),
actionBttn("runUMAP_xenium", "Run UMAP", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp3).removeClass('disabled')"),
),
br(),
br(),
column(12,
br(),
plotlyOutput("umap_xenium", width = "50%"),
),
br(),
column(6,
uiOutput("xenium.gene3.select"),
plotlyOutput("xenium_feature.plot", width = "100%"),
),
plotOutput("spatialumap_xenium", width = "50%"),
),
),
tabPanel("Clustering", value = "test_sp2",
tags$script(
'
var tab_sp2 = $(\'a[data-value="test_sp2"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Visium'",
column(3,
selectInput("scAnalysis_sp1",
label = "Analysis method",
choices = c("Seurat", "GraphST"),
selected = "Seurat"),
),
conditionalPanel(
condition = "input.scAnalysis_sp1 == 'Seurat'",
fluidRow(
column(3,
numericInput("clus.res_spatial",
label = "Resolution used",
value = 0.6,
min = 0.1,
step = 0.1)
),
column(3,
numericInput("dim.used_spatial",
label = "PC to use",
value = 10,
min = 1,
step = 1),
),
column(2,
br(),
actionBttn("findCluster_spatial", "Find Clusters", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp3).removeClass('disabled')"),
#textOutput("cluster1.done"),
)),
br(),
plotlyOutput("Cluster2DPlot_spatial", width = "50%"),
br(),
#br(),
plotOutput("SpatialDimPlot", width = "100%"),
br(),
textOutput("silhouette_seurat")
),
conditionalPanel(
condition = "input.scAnalysis_sp1 == 'GraphST'",
#column(4,
# fileInput("tpmFile_graphst",
# label = "Spaceranger output (Accepted Format: .h5)",
# accept = ".h5"),
#),
#column(3,
# shinyFiles::shinyDirButton(id = 'dir_graphst', label = "Path to spaceRanger output file", title = "Sheets Folder Selector"),
# verbatimTextOutput("dir_graphst", placeholder = TRUE),
#),
fluidRow(
column(3,
numericInput("cluster_graphst",
label = "Number of clusters",
value = 15,
min = 2,
step = 1)
),
column(3,
selectInput("cluster_method_graphst",
label = "Clustering method",
choices = c("mclust", "leiden", "louvain"),
selected = "mclust")
),
column(2,
br(),
actionBttn("findCluster_graphst", "Find Clusters", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp3).removeClass('disabled')"),
),
),
br(),
column(12,
plotOutput("Cluster2DPlot_graphst", width = "100%")
),
br(),
plotOutput("SpatialDimPlot_GraphST", width = "100%"),
br(),
textOutput("silhouette_graphst")
),
),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
fluidRow(
column(3,
numericInput("clus.res_xenium",
label = "Resolution used",
value = 0.3,
min = 0.1,
step = 0.1)
),
column(3,
selectInput("dim.used_xenium",
label = "Dimensions",
choices = c(10:50),
selected = 30),
),
column(3,
br(),
actionBttn("findCluster_xenium", "Find Clusters", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp3).removeClass('disabled')"),
br()
)),
br(),
plotlyOutput("Cluster2DPlot_xenium", width = "100%")
),
),
tabPanel("DEGs", value = "test_sp3",
tags$script(
'
var tab_sp3 = $(\'a[data-value="test_sp3"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_sp2 == 'Seurat'",
fluidRow(
column(3,
selectInput("deg_method_sp",
label = "Type of DEG analysis",
choices = c("Celltype specific", "Pairwise DEGs"),
selected = "Celltype specific"),
),
),
conditionalPanel(
condition = "input.deg_method_sp == 'Celltype specific'",
fluidRow(
column(3,
selectInput("deg_spatial",
label = "Group by",
choices = c("seurat_clusters"),
selected = "seurat_clusters"),
),
column(3, numericInput("min_pct_spatial",
label = "min.pct",
value = 0.75,
min = 0,
step = 0.01)
),
column(3, numericInput("logfc_spatial",
label = "logfc.threshold",
value = 0.75,
min = 0,
step = 0.01)
),
column(3, selectInput("test.use_spatial",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
),
column(4,
actionBttn("doDeg_spatial", "Run DEGs", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp4).removeClass('disabled')")
)),
br(),
column(12,
DT::dataTableOutput("Deg.table_spatial")
),
br(),
br(),
plotOutput("Deg.heatmap_spatial", width = "100%"),
),
conditionalPanel(
condition = "input.deg_method_sp == 'Pairwise DEGs'",
#br(),
column(12,
column(3,
selectInput("deg_sp1",
label = "Group by",
choices = c("seurat_clusters"),
selected = "seurat_clusters"),
),
column(3,
uiOutput("gene_sp.select"),
),
column(3,
uiOutput("gene_sp1.select"),
),
br(),
column(3,
actionBttn("doVolcano_sp", "Run Pairwise DEGs", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_sp4).removeClass('disabled')"),
#bsPopover("doVolcano", "Perform Pairwise DEG Analysis","Press to run Pairwise DEG Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
),
column(12,
column(3, numericInput("min_pct_sp1",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, numericInput("logfc_sp1",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
column(3, selectInput("test.use_sp1",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
)
),
column(12,
plotOutput("volcano.plot_sp1", width = "75%"),
),
br(),
column(12,
plotOutput("dega_sp.plot", width = "100%")
),
),
),
),
tabPanel("Deconvolution", value = "test_sp4",
tags$script(
'
var tab_sp4 = $(\'a[data-value="test_sp4"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
conditionalPanel(
condition = "input.scAnalysis_sp2 == 'Seurat' | input.scAnalysis_sp2 == 'GraphST'",
column(4,
br(),
actionBttn("loadexample_ref", "Load example reference dataset", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_sp5).removeClass('disabled')"),
),
tags$head(tags$script('
Shiny.addCustomMessageHandler("myCallbackHandler2",
function(typeMessage) {console.log(typeMessage)
if(typeMessage == 1){
console.log("got here");
$("a:contains(Single cell RNA-Sequencing)").click();
}
});
')),
column(4,
br(),
actionBttn("loaduser_ref", "Load and process user reference dataset", icon = icon("hand-o-right"), size = "sm"),
),
column(12,
#br(),
#actionButton("process_scRNA", "Process reference dataset", icon = icon("hand-pointer-o")),
br(),
br(),
plotOutput("scRNAPlot", width = "75%")),
column(9,
br(),
actionBttn("vis_spRNA", "Visualise spatial dataset", icon = icon("hand-pointer-o"), size = "sm"),
br(),
br(),
plotOutput("spRNAPlot", width = "100%")),
),
column(4,
br(),
selectInput("scAnalysis_sp2",
label = "Analysis method",
choices = c("Seurat", "GraphST"),
selected = "Seurat"),
),
conditionalPanel(
condition = "input.scAnalysis_sp2 == 'Seurat'",
column(4,
br(),
br(),
actionBttn("doDeconv", "Deconvolute", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp5).removeClass('disabled')"),
),
br(),
column(9,
uiOutput("ct.select"),
plotOutput("DeconvPlot", width = "100%")
),
),
conditionalPanel(
condition = "input.scAnalysis_sp2 == 'GraphST'",
column(4,
br(),
br(),
actionBttn("doDeconv_graphst", "Deconvolute", icon = icon("hand-pointer-o"), size = "sm", onclick = "$(tab_sp5).removeClass('disabled')"),
),
br(),
column(9,
uiOutput("graphst_ct.select"),
plotOutput("DeconvPlot_graphst", width = "100%")
),
)),
tabPanel("Data Visualization", value = "test_sp5",
tags$script(
'
var tab_sp5 = $(\'a[data-value="test_sp5"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
tabsetPanel(id="vis_sp",
tabPanel(title="Visualize genes", value="Vis_sp_panel1",
fluidRow(
column(3,
br(),
selectInput("deg_sp",
label = "Group by",
choices = c("seurat_clusters"),
selected = "seurat_clusters"),
),
column(6,
br(),
uiOutput("deg.gene.select_spatial"),
actionBttn("Vis_spatial", "Visualize", icon = icon("hand-pointer-o"), size = 'sm', onclick = "$(tab_sp5).removeClass('disabled')"),
plotlyOutput("Deg.plot_spatial", width = "150%"),
br(),
plotOutput("Deg1.plot_spatial", width = "150%"),
br(),
plotOutput("Deg2.plot_spatial", width = "150%"),
),
column(4,
downloadBttn('download_violn_spatial', 'Download Violin plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_feature_spatial', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
column(4,
downloadBttn('download_ridge_spatial', 'Download Feature plot (as png)', size = 'sm'),
br(),
),
)
),
tabPanel(title="Visualize Spatial Domains", value="Vis_sp_panel2",
conditionalPanel(
condition = "input.scAnalysis_sp1 == 'Seurat' & input.scAnalysis_platform == 'Visium'",
fluidRow(
column(6,
br(),
uiOutput("sp.cluster.select"),
plotOutput("sp1.plot", width = "100%"),
),
)),
conditionalPanel(
condition = "input.scAnalysis_sp1 == 'GraphST' & input.scAnalysis_platform == 'Visium'",
fluidRow(
column(6,
br(),
uiOutput("sp.cluster1.select"),
plotOutput("sp2.plot", width = "100%"),
),
)),
conditionalPanel(
condition = "input.scAnalysis_platform == 'Xenium'",
column(4,
actionBttn("Vis_sp_xenium", "Visualize", icon = icon("hand-pointer-o"), size = "sm")
),
fluidRow(
column(6,
uiOutput("xenium.cluster.select"),
plotOutput("xenium1.plot", width = "100%"),
),
)),
),
)
),
tabPanel("GSEA", value = "test_sp5",
tags$script(
'
var tab_sp5 = $(\'a[data-value="test_sp5"]\').parent().addClass("disabled");
$(function(){
$(tab_sp.parent()).on("click", "li.disabled", function(e) {
e.preventDefault();
return false;
});
});
'
),
column(12,
br(),
column(2, selectInput("species_gsea_sp",
label = "Species",
choices = c("Homo sapiens", "Mus musculus"))
),
column(2, selectInput("category_gsea_sp",
label = "Collection",
choices = c("H", "C2", "C5", "C7", "C8"))
),
column(2,
uiOutput("gsea.ct_sp1.select"),
),
column(2,
uiOutput("gsea.ct_sp2.select"),
),
column(2, numericInput("min_pct_sp1",
label = "min.pct",
value = 0.25,
min = 0,
step = 0.01)
),
column(2, numericInput("logfc_sp1",
label = "logfc.threshold",
value = 0.25,
min = 0,
step = 0.01)
),
),
column(12,
column(2, selectInput("test.use_sp1",
label = "Test use",
choices = c("wilcox", "bimod", "roc", "t", "negbinom", "poisson", "LR", "MAST", "DESeq2"))
),
column(4,
uiOutput("gsea_sp.select"),
),
br(),
column(3,
actionBttn("gsea_sp", "Run gene set enrichment analysis", icon = icon("hand-pointer-o"), size = 'sm'),
#bsPopover("gsea", "Perform Gene Set Enrichment Analysis","Press to run Gene Set Enrichment Analysis", placement = "bottom", trigger = "hover", options = NULL),
),
textOutput("gsea_sp.done"),
br()
),
br(),
column(7,
plotOutput("gsea_sp_plot", width = "100%"),
),
br(),
br(),
column(12,
plotOutput("gsea_sp_plot1", width = "100%"),
),
br(),
br(),
DT::dataTableOutput("gsea_sp.table"),
column(4,
downloadBttn('download_gsea_sp.table', 'Download GSEA Results (in csv)', size = 'sm'),
),
),
),
),
##---------------Single cell multiomics module-----------##
tabPanel("Single cell multiomics",
navlistPanel(widths=c(2,10),
tabPanel("Overview",
h2(p("Workflow for scMultiomics module")),
br(),
imageOutput("multiomics_image"),
),
tabPanel("Upload your data",
column(9,
column(5,
#h4('Load Data:'),
wellPanel(
conditionalPanel(
condition = "input.scAnalysis_type == 'CITE-seq'",
titlePanel(h4(p("Load example data"))),
actionBttn("loadexample_cite", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_mult).removeClass('disabled')"),
),
conditionalPanel(
condition = "input.scAnalysis_type == 'Multiome'",
titlePanel(h4(p("Load example data"))),
actionBttn("loadexample_multiome", "Load example and run", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_mult).removeClass('disabled')"),
),
titlePanel(h4(HTML("<b>Load your input data</b>"))),
br(),
conditionalPanel(
condition = "input.scAnalysis_type == 'CITE-seq'",
fileInput("tpmFiles2",
label = "Cellranger output (Accepted Format: .h5)",
accept = ".h5",
multiple = T)),
conditionalPanel(
condition = "input.scAnalysis_type == 'Multiome'",
fileInput(inputId = 'tpmFiles3',
label = "Cellranger output (Accepted Format: .h5)",
accept = ".h5")),
conditionalPanel(
condition = "input.scAnalysis_type == 'Multiome'",
shinyFiles::shinyFilesButton(id = 'dir_multi_atac', label = "Path to fragments file", title = "Sheets Folder Selector", multiple = T),
verbatimTextOutput("dir_multi_atac", placeholder = TRUE)
),
selectInput("scAnalysis_type",
label = "scAnalysis type",
choices = c("CITE-seq", "Multiome"),
selected = "H5"),
selectInput("scAnalysis_mult",
label = "Analysis method",
choices = c("Seurat", "MOFA+"),
selected = "Seurat"),
column(6,
numericInput(inputId = "min.genes2",
label = "Min. genes",
value = 200,
min = 1)
),
column(6,
numericInput(inputId = "min.cells2",
label = "Min. cells",
value = 3,
min = 1)
),
textInput(inputId = "projName2",
label = "Project Name",
value = "Multiomics"),
fluidRow(
actionBttn("loadButton2", "Load data", icon = icon("hand-o-right"), size = "sm", onclick = "$(tab_mult).removeClass('disabled')"),
actionBttn("reset_mult", "Reset", icon = icon("repeat"), size = "sm")
),
)),
chooseSliderSkin("Modern"),
titlePanel(h4(p("Quality control"))),
conditionalPanel(
condition = "input.scAnalysis_type == 'CITE-seq'",
column(6,
h4(p("RNA QC")),
plotOutput("nFeature_RNAPlot2a", width = "200%")
),
column(6,
h4(p("ADT QC")),
plotOutput("nFeature_RNAPlot2c", width = "150%")
),
),
conditionalPanel(
condition = "input.scAnalysis_type == 'Multiome'",
column(6,
h4(p("RNA QC")),
plotOutput("nFeature_RNAPlot2b", width = "200%")
),
column(6,
h4(p("ATAC QC")),
plotOutput("nFeature_RNAPlot2d", width = "150%")
),
),
conditionalPanel(
condition = "input.scAnalysis_type == 'CITE-seq'",
column(4,
numericInput("obsa",
label = "Min nFeature - RNA:",
value = 200,
min = 0,
step = 1),
),
column(4,
numericInput("obsa1",
label = "Max nFeature - RNA:",
value = 2500,
min = 0,
step = 1),
),
column(3,
numericInput("obsa2",
label = "Mt%:",
value = 5,
min = 0,
step = 1),
),
column(4,
numericInput("obsa3",
label = "Min nCount - ADT:",
value = 200,
min = 0,
step = 1),
),
column(4,
numericInput("obsa4",
label = "Max nCount - ADT:",
value = 50000,
min = 0,
step = 1),
),
),
conditionalPanel(
condition = "input.scAnalysis_type == 'Multiome'",
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment