Bio::DB::Fasta makeid sub

[mindlessbrain]

Hi there,

I am having trouble getting the makeid option work for my fasta db.

Here is the header of my the first sequence in my file.

>gi|149538895|ref|XP_001512649.1| PREDICTED: similar to Olfactory receptor 18 (Olfactory receptor 145-1) (Olfactory receptor TPCR34) [Ornithorhynchus anatinus]

What I want is to make the ids the gi number.

#!/usr/bin/perl

use Bio::SeqIO;
use Bio::Seq;
use Bio::DB::Fasta;
use Bio::PrimarySeq;
use Bio::LocationI;
use Bio::Tools::SeqWords;

sub make_my_id();

#Directory of fasta files
my $file = "/home/[user]/Desktop/Olfactory/Ornithorhynchus_anatinus_olfactory/ornithorhynchus_anatinus_sample.fasta";

######################## With Subroutine ##########################
my $db = Bio::DB::Fasta->new($file,-reindex, -makeid=>\&make_my_id);
 
my @names = $db->ids;
foreach (@names) {print $_, "\n";}

#my $seq = $db->seq('149538895'); 
#print $seq, "\n\n";

######Subroutines#################
sub make_my_id {
  my $line = shift;
  # get a different id from the fasta header, e.g.
  $line =~ m/(\|.*?\|)/;
  my $ID = $1;
  $ID = substr($ID, 1, -1);
  #print $ID, "\n";
  return $1;
}

Any insight as to whats going wrong would be very appreciated! Thanks!

[kmcarr]

Options for BioPerl objects need to be passed as key=>value pairs; simply having "-reindex" in your option list will not work. Change this to "-reindex=>1".

May I also suggest a simplification of your regex. If you know that the only files you will be parsing have NCBI formatted deflines then they will always start with ">gi|<gi_number>|" and gi_number is always an integer.

Replace:

$line =~ m/(\|.*?\|)/;
my $ID = $1;
$ID = substr($ID, 1, -1);

With:

my ($ID) = $line =~ m/^gi\|(\d+)\|/