Mercurial > repos > mvdbeek > r_goseq_1_22_0

comparison getgo.r @ 6:0e9424413ab0 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/ARTbio/tools-artbio/tree/master/tools/goseq_1_22_0 commit f95b47ed1a09ce14d3b565e8ea56d8bf12c35814-dirty
author mvdbeek
date Thu, 03 Mar 2016 09:56:51 -0500
parents b79c65c90744
children 04b9c519d3e1
comparison
equal deleted inserted replaced
5:b79c65c90744 6:0e9424413ab0
1 suppressWarnings(suppressMessages(library(goseq))) 1 options( show.error.messages=F, error = function () { cat( geterrmessage(), file=stderr() ); q( "no", 1, F ) } )
2 suppressWarnings(suppressMessages(library(optparse))) 2
3 suppressWarnings(suppressMessages(library(rtracklayer))) 3 # we need that to not crash galaxy with an UTF8 error on German LC settings.
4 suppressWarnings(suppressMessages(library(reshape2))) 4 loc <- Sys.setlocale("LC_MESSAGES", "en_US.UTF-8")
5
6 suppressPackageStartupMessages({
7 library("goseq")
8 library("optparse")
9 library("rtracklayer")
10 library("reshape2")
11 })
12
5 sink(stdout(), type = "message") 13 sink(stdout(), type = "message")
6 14
7 option_list <- list( 15 option_list <- list(
8 make_option(c("-gtf", "--gtf"), type="character", help = "Path to GTF file for which to fetch GO data"), 16 make_option(c("-gtf", "--gtf"), type="character", help = "Path to GTF file for which to fetch GO data"),
9 make_option(c("-g", "--genome"), type="character", help = "Genome [used for looking up GO categories]"), 17 make_option(c("-g", "--genome"), type="character", help = "Genome [used for looking up GO categories]"),
13 ) 21 )
14 22
15 parser <- OptionParser(usage = "%prog [options] file", option_list=option_list) 23 parser <- OptionParser(usage = "%prog [options] file", option_list=option_list)
16 args = parse_args(parser) 24 args = parse_args(parser)
17 25
18 # Vars: 26 # vars
19 27
20 gtf = args$gtf 28 gtf = args$gtf
21 genome = args$genome 29 genome = args$genome
22 gene_id = args$gene_id 30 gene_id = args$gene_id
23 output = args$output 31 output = args$output
24 cats = unlist(strsplit(args$cats, ',')) 32 cats = unlist(strsplit(args$cats, ','))
33
34 # retrieve and transform data
25 genes = unique(import.gff(gtf)$gene_id) 35 genes = unique(import.gff(gtf)$gene_id)
26 go_categories = getgo(genes, genome, id, fetch.cats=cats) 36 go_categories = getgo(genes, genome, gene_id, fetch.cats=cats)
27
28 # transform go category list to sth. more manipulatable in galaxy
29 go_categories <- lapply(go_categories, unlist)
30 go_categories = goseq:::reversemapping(go_categories) 37 go_categories = goseq:::reversemapping(go_categories)
31 go_categories = melt(go_categories) 38 go_categories = melt(go_categories)
32 colnames(go_categories) = c("#gene_id", "go_category")
33 39
34 write.table(go_categories, output, sep="\t", row.names = FALSE, quote = FALSE) 40 write.table(go_categories, output, sep="\t", col.names = FALSE, row.names = FALSE, quote = FALSE)
35 sessionInfo() 41 sessionInfo()