-
Notifications
You must be signed in to change notification settings - Fork 0
/
deseq_version_test.R
138 lines (128 loc) · 6.11 KB
/
deseq_version_test.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
if (interactive()){
basedir<<-file.path(Sys.getenv("CNA"),"rstudio")
} else {
basedir<<-"."
}
##================================================================================
outdir=file.path(basedir,"results")
annotdir = file.path(basedir,"info")
countdir=file.path(basedir,"counts")
dir.create(outdir)
##================================================================================
suppressPackageStartupMessages(library("optparse"))
option_list <- list(
make_option("--table", default=file.path(basedir,"info","calcineurin_sample_table.csv"),
help = "Sample table file for count data [default: \"%default\"]"),
make_option("--label", default="",
help = "Prefix LABEL to output file names"),
make_option("--fc", type="numeric", default=2,
help="Minimum fold change cutoff for genes"),
make_option("--fdr", type="numeric", default=0.2,
help="Maximum false discovery rate (FDR) cutoff for genes")
)
opt <- parse_args(OptionParser(option_list=option_list))
##================================================================================
library("DESeq2")
## library("RColorBrewer")
## library("gplots")
writeLines(capture.output(sessionInfo()), file.path(outdir,"sessionInfo.txt"))
##================================================================================
# counttab.file=file.path(basedir,"info","calcineurin_sample_table_drop_bad_cnako.csv")
counttab.file=opt$table
outbase = file.path(outdir,opt$label)
##================================================================================
sampleData = read.csv(counttab.file,comment.char="#", colClasses=c("character","numeric","character","factor","factor"))
sampleData$genotype = factor(sampleData$genotype, levels=c("WT", "KI_CNA1", "KI_CRZ1", "KO_cna1", "KO_crz1"))
sampleTable = transform(sampleData, sampleName=sample_name,fileName=sample_file,condition=genotype)
rownames(sampleTable) = sampleTable$sample_num
sampleTable <- subset(sampleTable, select = c(sampleName,fileName,condition,temp) )
ddsHTSeq <- DESeqDataSetFromHTSeqCount(sampleTable = sampleTable,
directory = countdir,
design= ~ condition)
## design(dds) <- formula(~ type + condition)
design(ddsHTSeq) <- formula(~ temp + condition)
ddsHTSeq <- DESeq(ddsHTSeq,betaPrior = FALSE)
res <- results(ddsHTSeq)
head(res)
print("----------------------------------------")
print("----------------------------------------\nresultsNames")
print("----------------------------------------")
print(resultsNames(ddsHTSeq))
resultsNames(ddsHTSeq)[3:6]
FindDiffGenes = function(ddsHTSeq,outbase, fdrcutoff=0.05, fccutoff=2,countfilter=FALSE){
## stop("need to do filtering for each comparison????")
log2fc = log2(fccutoff)
# for(var in seq) expr
ListOfGeneVecs = list()
for(sample in c("KI_CNA1","KI_CRZ1","KO_cna1","KO_crz1")) {
coeff = paste("condition",sample, "vs_WT",sep="_")
outfile = paste(sep="_",coeff, "results",fileend)
## if (packageVersion("DESeq2")=="1.2.10"){
## coeff = paste("condition",sample, "vs_WT",sep="_")
## outfile = paste(sep="_",coeff, "results",fileend)
## }else{
## coeff = paste("condition",sample,sep="")
## outfile = paste("condition",sample, "vs_WT","results",fileend,sep="_")
## }
print("coeff")
print(coeff)
cur.res = results(ddsHTSeq,name=coeff)
cur.res = cur.res[order(cur.res$padj),]
## print(sample)
print(
table(cur.res$padj < fdrcutoff,
abs(cur.res$log2FoldChange) >= log2fc,
dnn=c(paste("FDR<",fdrcutoff), paste("FC>",fccutoff)))
)
if (countfilter) {
print("Filtering Genes with mean counts less than 10 or NA pvalue")
use <- cur.res$baseMean >= 10 & !is.na(cur.res$pvalue)
table(use)
resFilt <- cur.res[use,]
resFilt$padj <- p.adjust(resFilt$pvalue, method="BH")
sum(cur.res$padj < .1, na.rm=TRUE)
sum(resFilt$padj < .1, na.rm=TRUE)
cur.res = resFilt
print(
table(cur.res$padj < fdrcutoff,
abs(cur.res$log2FoldChange) >= log2fc,
dnn=c(paste("FDR<",fdrcutoff), paste("FC>",fccutoff)))
)
filtered="_filt"
} else {
filtered=""
}
fileend=paste(fdrcutoff*100,"fdr_", fccutoff,"fc",filtered,".csv",sep="")
filt.res = cur.res[which((cur.res$padj < fdrcutoff) &
(abs(cur.res$log2FoldChange) >= log2fc)),]
outfile = paste("condition",sample, "vs_WT","results",fileend,sep="_")
write.csv(as.data.frame(filt.res),file=paste(sep="",outbase,outfile))
ListOfGeneVecs[[sample]] = row.names(filt.res)
}
KOcna1Genes = ListOfGeneVecs[["KO_cna1"]]
KOcrz1Genes = ListOfGeneVecs[["KO_crz1"]]
write.csv(KOcna1Genes,file=paste(sep="",outbase,paste("cna1ko_genes",fileend,sep="_")))
write.csv(KOcrz1Genes,file=paste(sep="",outbase,paste("crz1ko_genes",fileend,sep="_")))
## write.csv(intersect(KOcna1Genes,KOcrz1Genes),file=paste(sep="",outbase,paste("cna1ko_crz1ko_intersect",fileend,sep="_")))
## write.csv(setdiff(KOcna1Genes,KOcrz1Genes),file=paste(sep="",outbase,paste("cna1ko_unique",fileend,sep="_")))
## write.csv(setdiff(KOcrz1Genes,KOcna1Genes),file=paste(sep="",outbase,paste("crz1ko_unique",fileend,sep="_")))
## return(list("KOcna1Genes" = KOcna1Genes,"KOcrz1Genes"=KOcrz1Genes))
return(ListOfGeneVecs)
}
fclist = c(opt$fc)
fdrlist = c(opt$fdr)
##------------------------------------------------------------
filtlist = c(TRUE)
for (curfc in fclist) {
for (countfilter in filtlist){
for (curfdr in fdrlist){
if (countfilter) {
filtered="_filt"
} else {
filtered=""
}
fileend=paste(curfdr*100,"fdr_", curfc,"fc",filtered,sep="")
fdg = FindDiffGenes(ddsHTSeq, outbase,fdrcutoff=curfdr,fccutoff=curfc,countfilter=countfilter)
}
}
}