bed2sort

#!/bin/bash
#PBS -l nodes=1:ppn=4

#### usage ####
usage() {
	echo Program: "bed2sort (sort input BED file by width of region)"
	echo Author: BRIC, University of Copenhagen, Denmark
	echo Version: 1.0
	echo Contact: pundhir@binf.ku.dk
	echo "Usage: bed2sort -i <file>"
	echo "Options:"
	echo " -i <file>   [input file containing genomic coordinate in BED format (can be stdin)]"
	echo " -h          [help]"
	echo
	exit 0
}

#### parse options ####
while getopts i:h ARG; do
	case "$ARG" in
		i) INPUTBEDFILE=$OPTARG;;
		h) HELP=1;;
	esac
done

## usage, if necessary file and directories are given/exist
if [ -z "$INPUTBEDFILE" -o "$HELP" ]; then
	usage
fi

## create temporary BED file if input is from stdin
if [ "$INPUTBEDFILE" == "stdin" ]; then
    TMP=$(cat /dev/urandom | tr -dc 'a-zA-Z0-9' | fold -w 32 | head -n 1)
    while read LINE; do
        echo ${LINE}
    done | perl -ane '$line=""; foreach(@F) { $line.="$_\t"; } $line=~s/\t$//g; print "$line\n";' > $TMP
    INPUTBEDFILE=$TMP
fi

## sort input bed file by width
NCOL=$(head -n 1 $INPUTBEDFILE | perl -ane 'printf("%d", scalar(@F));')
NEW_COL=$(head -n 1 $INPUTBEDFILE | perl -ane 'printf("%d", scalar(@F)+1);')
less $INPUTBEDFILE | perl -ane '$len=$F[2]-$F[1]; chomp($_); print "$_\t$len\n";' | sort -k $NEW_COL"n",$NEW_COL | cut -f 1-$NCOL