/DeepTransferLearning

Codes for the deep transfer learning in bioinformatics

Normal RoadMap Data

Codes applied for the roadmap data analysis part.

Downsamling & Creating sorted & indexed bam files from tagAlign files

If you see '-bash: shuf: command not found' run following code:

brew install coreutils
for tg in *.tagAlign.gz
do 
	zless $tg | shuf -n 15000000 | bedtobam -i - -g hg19.txt | samtools sort -o "${tg%%.*}".sorted.bam 
	samtools index "${tg%%.*}".sorted.bam
done

Check the number of Reads for each bam file

for file in *.sorted.bam; do samtools idxstats "${file%%.*}".sorted.bam | cut -f3 | awk 'BEGIN {total=0} {total += $1} END {print total}'; done

Counting the Number of Reads around TSS

require(TxDb.Hsapiens.UCSC.hg19.knownGene)
# To avoid have to type the whole package name every time, we use the variable name txdb
txdb <- TxDb.Hsapiens.UCSC.hg19.knownGene


pms <- promoters(genes(txdb), upstream = 5000, downstream = 5000)


gr<- unlist(tile(pms, width=100))


df <- data.frame(seqnames=seqnames(gr),
                 starts=start(gr),
                 ends=end(gr),
                 names=c(rep(".", length(gr))),
                 scores=c(rep(".", length(gr))),
                 strands=strand(gr))

write.table(df, file="foo.bed", quote=F, sep="\t", row.names=F, col.names=F)