vsbuffalo

# A quick function to save a PBM (http://en.wikipedia.org/wiki/Netpbm_format)
# visualize *a lot* of missing data pretty quickly (outside of R).

writeMissingPBM <- function(x, file) {
  dims <- dim(x)
  x[] <- as.integer(is.na(x))
  con <- file(file, open="wt")
  writeLines(sprintf("P1\n%d %d", ncol(x), nrow(x)), con)
  write.table(x, file=con, sep=" ", col.names=FALSE, row.names=FALSE, quote=FALSE)
  close(con)

arunsrinivasan

require(reshape2)

# data.table commit (1048)
require(data.table)
# Loading required package: data.table
# data.table 1.8.11  For help type: help("data.table")

set.seed(1)
N <- 2e7 # size of DT

arq5x

export SAMPLES="2484-AJ-0001 2484-AJ-0002 2484-AJ-0003"

######################################
# Make FASTQ
######################################
export OVHOME=/home/arq5x/cphg-home/cphg-quinlan/projects/ov-cell-lines
export STEPNAME=ovc-fastq
for sample in `echo $SAMPLES`
do
export QSUB="qsub -W group_list=cphg_arq5x -q arq5xlab -V -l select=1:mem=8000m:ncpus=1 -N $STEPNAME -m bea -M arq5x@virginia.edu";

stephenturner

## devtools::install_github("stephenturner/msigdf")
library(msigdf)
library(dplyr)
library(clusterProfiler)

c2 <- msigdf.human %>% 
    filter(collection == "c2") %>% select(geneset, entrez) %>% as.data.frame

data(geneList)
de <- names(geneList)[1:100]

nickloman

rsync -av rsync://ftp.ncbi.nlm.nih.gov/genomes/Bacteria --include "*/" --include "Bacteria/Escherichia*/*.fna" --exclude=* .

# bonus script - concatenate chromosomes and plasmids into single fasta file, make sure the files don't already exist

find . -mindepth 1 -type d | xargs -L 1 -I '{}' find {} -name "*.fna" | while read i ; do cat "$i" >> `dirname "$i"`.fasta ; done

jennybc

What if a function in a package has no examples? Or is poorly exampled? Wouldn't it be nice to find functioning instances of it "in the wild"?

Via Twitter, Noam Ross taught me a clever way to do such searches on GitHub. Put this into the GitHub search box to see people using the llply() function from plyr:

"llply" user:cran language:R

Or just click here.

seandavi

idat2lumibatch <- function(filenames) {
  # filenames is a character vector of iDAT filenames
  require(illuminaio)
  require(lumi)
  idatlist = lapply(filenames,readIDAT)
  exprs = sapply(idatlist,function(x) {
    return(x$Quants$MeanBinData)})
  se.exprs = sapply(idatlist,function(x) {
    return(x$Quants$DevBinData/sqrt(x$Quants$NumGoodBeadsBinData))})
  beadNum = sapply(idatlist,function(x) {

stephenturner

## See http://gettinggeneticsdone.blogspot.com/2013/06/customize-rprofile.html

## Load packages
library(BiocInstaller)

## Don't show those silly significanct stars
options(show.signif.stars=FALSE)

## Do you want to automatically convert strings to factor variables in a data.frame?
## WARNING!!! This makes your code less portable/reproducible.

stephenturner

# Compile and install htslib, bcftools, samtools 1.9 to home directory

# Set up dirs where stuff will be installed
mkdir -p ${HOME}/bin/htslib
mkdir -p ${HOME}/bin/bcftools
mkdir -p ${HOME}/bin/samtools

# Make a dir to build
mkdir -p ${HOME}/tmp

stephenturner

Installing gcc 4.8 and Linuxbrew on CentOS 6

The GCC distributed with CentOS 6 is 4.4.7, which is pretty outdated. I'd like to use gcc 4.8+. Also, when trying to install Linuxbrew you run into a dependency loop where Homebrew's gcc depends on zlib, which depends on gcc. Here's how I solved the problem.

Note: Requires sudo privileges.

Resources:

• http://superuser.com/a/676337/88393: Forum response on using CERN's open Scientific Linux distribution of RHEL's developer toolset.
• http://linux.web.cern.ch/linux/devtoolset/: CERN's developer toolset installation instructions.