RNAseqDownstream2report/RNAseq_6_enrichfunc.R

#!/usr/bin/env Rscript
###Copyright 2019 Ying Yu from Fudan-PGx group 
# example:
# Rscript RNAseq_6_enrichfunc.R -i rnaseq_degs_acrossgroups.csv

suppressPackageStartupMessages(library("optparse"))
suppressPackageStartupMessages(library("clusterProfiler"))

# specify our desired options in a list
# by default OptionParser will add an help option equivalent to 
# make_option(c("-h", "--help"), action="store_true", default=FALSE, 
#               help="Show this help message and exit")

# input input list , rds, from * to *

option_list <- list( 
    make_option(c("-o", "--out_dir"), type="character",default="./",
        help="The output directory [default ./]"),
    make_option(c("-i", "--input"),type="character", default=NULL,
        help="The input DEG list in csv format. The first column: gene; second column: group. Required! "),
	make_option(c("-e", "--type_gene_id"),type="character", default="EnsemblGID",
        help="The type of gene symbol. Could be either of EnsemblGID/EntrezID/GeneSymbol [default: EnsemblGID]"),
	make_option(c("-f", "--pvalueCutoff"), type="double",default=0.05,metavar="number",
		help="Cutoff value of p value. [default: 0.05]"),
	make_option(c("-m", "--pAdjustMethod"), type="character",default="BH",
		help="Method of adjust p value. One of \"holm\", \"hochberg\", \"hommel\", \"bonferroni\", \"BH\", \"BY\", \"fdr\", \"none\". [default: BH]"),
	make_option(c("-q", "--qvalueCutoff"), type="double",default=0.2,metavar="number",
		help="Cutoff value of q value. [default: 0.2]"),
   	make_option(c("-p", "--project_code"), type="character",default="rnaseq",
        help="Project code, which is used as prefix of output file. [default: rnaseq]"),
	make_option(c("-d", "--ref_rdata_dir"), type="character",default="./",
        help="The directory of reference files: human_c2_v5p2.rdata, human_c5_v5p2.rdata and ID_convert_table.rds. [default: ./]")
		)


# get command line options, if help option encountered print help and exit,
# otherwise if options not found on command line then set defaults, 
opt <- parse_args(OptionParser(option_list=option_list))

if (is.null(opt$input)){
  print_help(opt_parser)
  stop("At least one argument must be supplied (input file).", call.=FALSE)
}

message("Need to connected to the Internet.")

##import file
out_dir<-paste(gsub("/$","",opt$out_dir),"/",sep="")
gene<-read.csv(opt$input,header=T,stringsAsFactors=F)

##########################
#########ID convert#######
##########################

message("Begin ID conversion.")

refdir<-paste(gsub("/$","",opt$ref_rdata_dir),"/",sep="")

if(length(grep("ID_convert_table.rds",dir(refdir)))>0){
idconvert<-readRDS(paste(refdir,"ID_convert_table.rds",sep=""))
}else{
 stop("Cannot find ID_convert_table.rds in the ref_rdata_dir. Exit!", call.=FALSE)
}

if(opt$type_gene_id=="EnsemblGID"){
gene$EntrezID<-idconvert$EntrezID[match(gene[,1],idconvert$EnsemblID)]
if(length(which(is.na(gene$EntrezID)))==nrow(gene)){
 stop("Cannot convert Ensembl gene ID to Entrez gene ID. Exit!", call.=FALSE)
}
}

if(opt$type_gene_id=="GeneSymbol"){
gene$EntrezID<-idconvert$EntrezID[match(gene[,1],idconvert$GeneSymbol)]
if(length(which(is.na(gene$EntrezID)))==nrow(gene)){
 stop("Cannot convert GeneSymbol to Entrez gene ID. Exit!", call.=FALSE)
}
}

if(opt$type_gene_id=="EntrezID"){
gene$EntrezID<-gene[,1]
}

message("Finish ID conversion.")

##########################
#########Enrich GO#######
##########################

groupn<-unique(gene[,2])
if(length(groupn)==0){
 message("Warning: no group infomation. Function enrichment will be conducted as one group.")
}else{
message(paste("A number of ", length(groupn)," group(s) is detected. Function enrichment will be conducted in ",length(groupn), " group(s).",sep=""))
}

####
egoall<-c()
ekeggall<-c()
if(length(groupn)==0){
g1<-gene$EntrezID
g1<-g1[!g1==""]
#conduct enrichment

message("Contucting enrichment analysis on GO terms...")
ego<-data.frame(enrichGO(g1, 'org.Hs.eg.db', ont = 'ALL', pvalueCutoff = opt$pvalueCutoff, pAdjustMethod = opt$pAdjustMethod, qvalueCutoff = opt$qvalueCutoff))

message("Contucting enrichment analysis on KEGG pathways...")
ekg<- data.frame( enrichKEGG(g1, organism = "hsa", keyType = "kegg", pvalueCutoff = opt$pvalueCutoff, pAdjustMethod = opt$pAdjustMethod, qvalueCutoff = opt$qvalueCutoff))
  
#modify output
if(!nrow(ego)==0){
ego1<-cbind(groupn[i],ego)
colnames(ego1)[1]<-c("versus")
egoall<-rbind(egoall,ego1)
}

if(!nrow(ekg)==0){
ekg1<-cbind(groupn[i],ekg)
colnames(ekg1)[1]<-c("versus")
ekeggall<-rbind(ekeggall,ekg1)
}

}else{

for (i in 1:length(groupn)){

message(paste("Group ", groupn[i],sep=""))

g1<-gene$EntrezID[gene[,2]==groupn[i]]
g1<-g1[!g1==""]
#conduct enrichment

message("Contucting enrichment analysis on GO terms...")
ego<-data.frame(enrichGO(g1, 'org.Hs.eg.db', ont = 'ALL', pvalueCutoff = opt$pvalueCutoff, pAdjustMethod = opt$pAdjustMethod, qvalueCutoff = opt$qvalueCutoff))

message("Contucting enrichment analysis on KEGG pathways...")
ekg<- data.frame( enrichKEGG(g1, organism = "hsa", keyType = "kegg", pvalueCutoff = opt$pvalueCutoff, pAdjustMethod = opt$pAdjustMethod, qvalueCutoff = opt$qvalueCutoff))

  if(!nrow(ego)==0){
ego1<-cbind(groupn[i],ego)
colnames(ego1)[1]<-c("versus")
egoall<-rbind(egoall,ego1)
message(paste(nrow(ego),"significant GO term(s) is(are) identified."))
}else{
message("No significant GO term is identified.")
}

if(!nrow(ekg)==0){
ekg1<-cbind(groupn[i],ekg)
colnames(ekg1)[1]<-c("versus")
ekeggall<-rbind(ekeggall,ekg1)
message(paste(nrow(ekg),"significant KEGG pathway(s) is(are) identified."))
}else{
message("No significant KEGG pathway is identified.")
}
}
}
message("Wrinting output...")

#write output
if(nrow(egoall)==0){
message("No significant GO term is identified across all tested groups.")
}else{
message(paste(nrow(egoall),"significant GO term(s) is(are) identified across all tested groups."))
rownames(egoall)<-c(1:nrow(egoall))
egoall$pvalue<-signif(egoall$pvalue,4)
egoall$p.adjust<-signif(egoall$p.adjust,4)
egoall$qvalue<-signif(egoall$qvalue,4)
write.csv(egoall,paste(out_dir,opt$project_code,"_GOenrich.csv",sep=""))
}

if(nrow(ekeggall)==0){
message("No significant KEGG pathway is identified across all tested groups.")
}else{
message(paste(nrow(ekeggall),"significant KEGG pathway(s) is(are) identified across all tested groups."))
rownames(ekeggall)<-c(1:nrow(ekeggall))
ekeggall$pvalue<-signif(ekeggall$pvalue,4)
ekeggall$p.adjust<-signif(ekeggall$p.adjust,4)
ekeggall$qvalue<-signif(ekeggall$qvalue,4)
write.csv(ekeggall,paste(out_dir,opt$project_code,"_KEGGenrich.csv",sep=""))
}
########