shuffleSequences_fastq.pl

[mmmm]

I am using the perl script (shuffleSequences_fastq.pl) to prepare the intervaled fastq file for velvet and I got the first read in read1 followed by the first read in read2 BUT then I got 3 different reads of reads 2 followed by their correspondence in read1 and so on- I was wondering if this ok or is this below script right?
a read1
a read2
b read1
b read2
c read2
d read2
c read1
d read1 and so on

shuffleSequences_fastq.pl

#!/usr/bin/perl

if (!@ARGV) {
print "Usage: $0 forward_reads.fa reverse_reaads.fa outfile.fa\n";
print "\tforward_reads.fa / reverse_reads.fa : paired reads to be merged\n";
print "\toutfile.fa : outfile to be created\n";
system.exit(0);
}

$filenameA = $ARGV[0];
$filenameB = $ARGV[1];
$filenameOut = $ARGV[2];

die "Could not open $filenameA" unless (-e $filenameA);
die "Could not open $filenameB" unless (-e $filenameB);

open FILEA, "< $filenameA";
open FILEB, "< $filenameB";

open OUTFILE, "> $filenameOut";

my ($lineA, $lineB);

$lineA = <FILEA>;
$lineB = <FILEB>;

while(defined $lineA) {
print OUTFILE $lineA;
$lineA = <FILEA>;
while (defined $lineA && $lineA !~ m/>/) {
print OUTFILE $lineA;
$lineA = <FILEA>;
}

print OUTFILE $lineB;
$lineB = <FILEB>;
while (defined $lineB && $lineB !~ m/>/) {
print OUTFILE $lineB;
$lineB = <FILEB>;
}
}

[mastal]

If you use a recent version of velvet, you don't need to use the shuffleSequences scripts, you can leave the reads as 2 separate files and use the flag '-separate' when running velveth.

I have used the shuffleSequences_fastq.pl script that comes with velvet in the past, and never had any problems.

It is not OK if your files don't interleave properly, you will have problems with the assembly or alignment programs further downstream.

[mastal]

Are you trying to interleave fasta files or fastq files?

If it is fastq files, your script won't work because the header lines for each read should begin with '@'.

Change the two lines with the regular expressions.

Change:

while (defined $lineA && $lineA !~ m/>/) {

to

while (defined $lineA && $lineA !~ m/^@/) {

[mmmm]

thank you very much- also, I should change
while (defined $lineB && $lineB !~ m/>/)
to
while (defined $lineB && $lineB !~ m/^@/)

[SES]

thank you very much- also, I should change
while (defined $lineB && $lineB !~ m/>/)
to
while (defined $lineB && $lineB !~ m/^@/)

I wouldn't edit the script for fasta files, instead just use the script designed for fastq. You can find the script on the velvet github site. This assumes 4 line fastq. You can try Pairfq (documentation) for a more flexible solution (reading multiline fasta/q, compressed/uncompressed), but you may want to start with trying the script distributed with velvet.