How To Transform An Interleaved Fasta File To A Sequential Fasta File

Question

[samlambrechts299]

Hi everybody,

I am trying to convert the sequences in a fasta file from the interleaved format to a sequential format

My input:

>gi|161085638|dbj|AB305033.1| 
        ATATGCCTGAAAGTGGCGGACGGGTGAGTAACACGTGGGTGACCTGCCTCGGAGTGGGGGATAACCATGG
        GAAACTGTGGCTAATACCGCATGGGCTTGTTGGCTTTGGCGGCCAACGAGTAAAGCTTTAGTGCTTCGAG
        AGGGGCCTGCGTCCGATTAGGTAGTTGGTGAGGTAATGGCTCACCAAGCCGATGATCGGTAGCTGGTCTG
>gi|161085638|dbj|AB305644.1| 
        ATATGCCTGAAAGTGGCGGACGGGTGAGTAACACGTGGGTGACCTGCCTCGGAGTGGGGGATAACCATGG
        GAAACTGTGGCTAATACCGCATGGGCTTGTTGGCTTTGGCGGCCAACGAGTAAAGCTTTAGTGCTTCGAG
        AGGGGCCTGCGTCCGATTAGGTAGTTGGTGAGGTAATGGCTCACCAAGCCGATGATCGGTAGCTGGTCTG

Desired output:

 >gi|161085638|dbj|AB305033.1| 
    ATATGCCTGAAAGTGGCGGACGGGTGAGTAACACGTGGGTGACCTGCCTCGGAGTGGGGGATAACCATGGGAAACTGCGGCCAACGAGTAAAGCTTTAGTGCTTC...
 >gi|161085638|dbj|AB305644.1| 
    ATATGCCTGAAAGTGGCGGACGGGTGAGTAACACGTGGGTGACCTGCCTCGGAGTGGGGGATAACCATGGGGCTAATACCGCATGGGCTTGTTGGCTTTGGCGGC...

After unsuccessfully trying to compile a script myself, I have found the following on http://phototrophic.net/node/37:

#!/usr/bin/perl                                                                                                                                                                     
$in = open(IN,"<file.fasta");while ($in=<IN>){chomp $in;if ($in=~m/^>/) { print "\n",$in,"\n";}else{print $in;}}

But when I try to use this I get bashed:

bash: syntax error near unexpected token `in'

If anyone can provide an answer why I get this syntax error, or can help me with a script to convert interleaved files to sequential files, that would be greatly appreciated

Best,

Sam

Answer 1

[SES]

There are a number of things wrong with that Perl code. You have to be careful taking code samples from the internet because sometimes you find really useful information on blogs, and other times you find code like this. I don't know why something is trying to interpret it as a Bash script but it could be multiple things. For just removing the line-wrapping you can use a BioPerl one-liner:

perl -MBio::SeqIO -e 'my $seqin = Bio::SeqIO->new(-fh => \*STDIN, -format => 'fasta'); while (my $seq = $seqin->next_seq) { print ">",$seq->id,"\n",$seq->seq,"\n"; }' < seqs.fasta > seqs_nowrap.fasta

I'll bet you can also do this with seqret or some other program from EMBOSS.

[samlambrechts299]

Thanks, will try this

[Kenosis]

Excellent use of Bio::SeqIO.

Answer 2

[Ashutosh Pandey]

See this other question on Biostar:

multiline fasta to single line fasta

[samlambrechts299]

aha ok, that explains why I didn't find an existing question/topic covering this problem, a matter of vocabulary. Thank you for posting this!

Answer 3

[Kenosis]

Here's another option:

use strict;
use warnings;

$/ = '>';
while (<>) {
    chomp;
    s/(.+?\n)(.+)/my $x = $2; $x =~ s|\s+||g; $_ = $x/se or next;
    print ">$1   $_\n";
}

Usage: perl script.pl inFile.fasta [>outFile.fasta]

The second, optional parameter will direct the output to a file.

This uses ">" as the fasta record seperator. A regex captures the id and seq, and removes whitespaces within the seq. All is then finally printed.

Hope this helps!