diff scripts/pickUniqPairFastq.pl.orig @ 21:9672fe07a232 draft default tip

planemo upload for repository https://github.com/portiahollyoak/Tools commit 0fea84d05f8976b8360a8b4943ecb01b87e3ade0-dirty
author mvdbeek
date Mon, 05 Dec 2016 09:58:47 -0500
parents
children
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/scripts/pickUniqPairFastq.pl.orig	Mon Dec 05 09:58:47 2016 -0500
@@ -0,0 +1,45 @@
+#!/share/bin/perl
+use Bio::Seq;
+
+die "perl $0 <sam> <output prefix>\n" if @ARGV<1;
+
+open m1,">$ARGV[1].1.fastq";
+open m2,">$ARGV[1].2.fastq";
+
+open in,$ARGV[0];
+my %pe;
+while(<in>)
+{
+	chomp;
+	my @f=split/\t/,$_,12;
+	## read number 1 or 2
+	my ($rnum)=$f[1]=~/(\d)$/;
+
+	## XT:A:* 
+	my ($xt)=$f[11]=~/XT:A:(.)/;
+
+	## revcom the read mapped to the reverse strand
+	if($f[1]=~/r/)
+	{
+		my $seq=Bio::Seq->new(-seq=>$f[9], -alphabet => 'dna');
+		$f[9]=$seq->revcom->seq;
+		$f[10]=reverse $f[10];
+	}
+	if (($rnum == 1) || ($rnum == 2))
+	{
+	    ${$pe{$f[0]}}[$rnum-1]=[$xt,$f[9],$f[10]];
+	}
+}
+close in;
+
+foreach my $id (keys %pe)
+{
+	my @rid=@{$pe{$id}};
+	if (($rid[0][1] ne "") && ($rid[1][1] ne "") && (($rid[0][0] eq "U" || $rid[1][0] eq "U")))
+	{
+		print m2 "@"."$id/2","\n",$rid[1][1],"\n","+$id/2","\n",$rid[1][2],"\n";
+		print m1 "@"."$id/1","\n",$rid[0][1],"\n","+$id/1","\n",$rid[0][2],"\n";
+	}
+}
+close m1;
+close m2;