Created
November 8, 2013 22:54
-
-
Save vsbuffalo/7378986 to your computer and use it in GitHub Desktop.
Example of GenomicRanges's tileGenome, which I think demonstrates its power. This might be a bit faster as a custom script in Python or C, but (1) this would take longer and (2) this is much more interactive (3) on real data, it's actually pretty fast. Stuff like this is why Bioconductor should be in every bioinformatician's toolkit.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
library(GenomicRanges) | |
summarizeByTile <- | |
# given a GRanges (or some sort of ranged data) object `x`, and a | |
# *corresponding* vector values to summarize `y` (these *must* | |
# correspond), calculate the summary per tile with the function `fun`. | |
# Note: this is still beta; wider tests coming, use with caution. | |
function(x, y, tiles, fun, mcol_name="y") { | |
stopifnot(length(x) == length(y)) | |
if (is(tiles, "GRangesList")) | |
tiles <- unlist(tiles) | |
# find overlaps between ranges object x and tiles | |
hits <- findOverlaps(x, tiles) | |
message(sprintf("summarizeByTile: %d overlaps found between x (%d rows) and genome tiles (%d rows)", | |
length(hits), length(x), length(tiles))) | |
# split y by hits in tiles | |
split_y <- split(y[queryHits(hits)], subjectHits(hits)) | |
if (!is.null(mcol_name)) { | |
tile_summaries <- tiles[unique(subjectHits(hits))] | |
mcols(tile_summaries)[mcol_name] <- sapply(split_y, fun) | |
return(tile_summaries) | |
} else | |
sapply(split_y, fun) | |
} | |
# toy example | |
chr1_length <- c(chr1=1000) | |
tiles <- tileGenome(chr1_length, tilewidth=100) | |
snps <- GRanges("chr1", IRanges(sample(seq_len(chr1_length)), width=1), genotype=sample(c("0/0", "0/1", "1/1"), chr1_length, replace=TRUE)) | |
tile_hets <- summarizeByTile(snps, mcols(snps)$genotype, tiles, fun=function(x) sum("0/1" == x)/length(x)) | |
## > tile_hets | |
## > summarizeByTile(snps, mcols(snps)$genotype, tiles, fun=function(x) sum("0/1" == x)/length(x)) | |
## summarizeByTile: 1000 overlaps found between x (1000 rows) and genome tiles (10 rows) | |
## GRanges with 10 ranges and 1 metadata column: | |
## seqnames ranges strand | y | |
## <Rle> <IRanges> <Rle> | <numeric> | |
## [1] chr1 [101, 200] * | 0.31 | |
## [2] chr1 [801, 900] * | 0.31 | |
## [3] chr1 [901, 1000] * | 0.43 | |
## [4] chr1 [401, 500] * | 0.3 | |
## [5] chr1 [601, 700] * | 0.42 | |
## [6] chr1 [701, 800] * | 0.31 | |
## [7] chr1 [201, 300] * | 0.28 | |
## [8] chr1 [501, 600] * | 0.34 | |
## [9] chr1 [301, 400] * | 0.28 | |
## [10] chr1 [ 1, 100] * | 0.35 | |
## --- | |
## seqlengths: | |
## chr1 | |
## 1000 | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment