http://qs321.pair.com?node_id=913678

garyboyd has asked for the wisdom of the Perl Monks concerning the following question:

Hi monks, I have a list of 121,000 fasta headers in a .txt file and I want to compare this to a multiple FASTA file containing 132,000 sequences. I have written a script to do this but it is too slow!!! The script puts all the lines from the .txt file in a hash and then compares the header of each FASTA sequence in the multiple FASTA file with the keys in the hash and if they match prints out a new FASTA sequence with the renamed header from the .txt file. Is there anything sensible I can do to speed this up?

Input txt file looks like:

F0Z7V0F01A03EB_210 F0Z7V0F01A03EB_180 F0Z7V0F01A03EB_136 F0Z7V0F01A03EB_362 F0Z7V0F01A03GP_298 F0Z7V0F01A03GP_205 etc...........

Input FASTA looks like:

>GJKKTUG01DYDGC GGGTATTCCTTCTCCACCTTGCAGCTAACATCAGTGTTTCGTCTACTCAAGCACGCCAAC ACGCCCTAGAGCGCCCTGTCCAGGGGATGGCAACCAACTCTGACCCTGCAAGTGCAGCAG ACATGAGGAATACAAACTACAATCTTTTACTTGATGATGCAATGCCGGACAAACTCTAGA >F0Z7V0F01EDB3V AAGGCGAGNGGTATCACGCAGTAAGTTACGGTTTTCGGGTAACGCGTCNGNGGNACTAAC CCACGGNGGGTAACCCGTCNCTACCGGTATAGGACTAAGGTTACCGGAACGTCGTGGGGT ACCCCCCGGACGGGGACCGTCCCCTCATANAGTCAACNGTNTGAGATGGACTAACTCAAA CCTAGTTTCAAGTACTATTTAACTTACTTACGTTACCCGTAATTTCGGCGTTTAGAGGCG etc....................

My script which is working VERY slowly is :

#!/usr/bin/perl use strict; use warnings; use Bio::SeqIO; use Data::Dumper; my %seq_id; my $fasta_id; open HEADER , "<FASTA.headers" or die $!; while (<HEADER>){ chomp $_; #print $_."\n"; $fasta_id = $_; $fasta_id =~ s/_.*//g ; #print $fasta_id."\n"; %seq_id = ("$fasta_id" => "$_"); # print Dumper (\%seq_id); my $infile=$ARGV[0] || die ("Please give me an input fasta + file\n"); my $inseq = new Bio::SeqIO(-format => 'fasta', -file => $infile); while (my $seq_obj = $inseq->next_seq ) { my $id = $seq_obj->id ; chomp $id; # print $id."\n"; my $seq = $seq_obj->seq ; if (exists ($seq_id{$id})) { print ">"; print $seq_id{$fasta_id}; print "\n".$seq."\n"; } } }