merge.R

#!/home/jhamp/.conda/envs/tfbs/bin/Rscript
args=commandArgs(TRUE)
min=as.numeric(args[1])
max=as.numeric(args[2])
folder=args[3]
splitted = read.table(paste(folder, "/pass2Tr.bed", sep=''), header=FALSE)
colnames(splitted) = c("chromosome", "start", "stop", "id", "score", "maxpos", "length")
p1 = read.table(paste(folder, "/pass1Tr.bed", sep=''), header=FALSE)
colnames(p1) = c("chromosome", "start", "stop", "id", "score", "maxpos", "length")
p1$maxpos = p1$start + p1$maxpos

splitted=rbind(splitted, p1)

splitted=splitted[which(splitted$stop - splitted$start >= min),]
splitted=splitted[which(splitted$stop - splitted$start <= max),]
splitted$id=make.unique(as.character(splitted$id))
splitted$length=splitted$stop - splitted$start

splitted=cbind(splitted, containsMaxpos=0)
splitted$containsMaxpos[intersect(which(splitted$start <= splitted$maxpos), which(splitted$stop > splitted$maxpos))] = 1
splitted$maxpos = splitted$maxpos - splitted$start
write.table(splitted, paste(folder, "/merged.bed", sep=''), row.names=FALSE, col.names=FALSE, quote=FALSE, sep='\t')