array of hashes in perl

Question

[natasha.sernova]

Dear collegues,
I need your advice. I have a mixture of protein sequences,
about a thousand. They are in fasta-format - their names are
always different, but their sequences sometimes are the same.
I would like to get rid of any repeats automatically.
Is there any simple way to do it? Hashes "seen" may help, but I don't know exactly how many hashes I will need to create.
To create the array of these hashes is too complicated to my mind.
Is it the only way to do it?
Many thanks for your help!

[Prakki Rama]

check How To Remove The Same Sequences In The Fasta Files? thread to remove duplicate sequences.

[natasha.sernova]

Many thanks! I am very bad in python, but this is a good reason to study it better.

I have to try, I have no choice.

[Prakki Rama]

Check pierre's post in the above link, where you can just run the command to remove duplicate without running any program.

[natasha.sernova]

Really? That's great! But I didn't quite understand, what "pierre's post"

you are talking about. It's my dream - just run the command to remove duplicates without running any program. Please, help me to find it! THOUSAND THANKS!

Natasha

[Prakki Rama]

:) Click this

[Neilfws]

I assume that by "repeats" you mean "duplicate sequences" as opposed to sequence repeats.

[natasha.sernova]

Yes, exactly. I mean multiple duplicate sequences.

Answer 1

[JC]

Well, you need to use the sequence as the "key" in the hash and the sequence_id as the "value", for example:

#!/usr/bin/perl
use strict;
use warnings;
my %seqs;
$/ = "\n>";
while (<>) {
    s/>//g;
    my ($id, @seq) = split (/\n/, $_);
    my $seq = join "", @seq;
    $seqs{$seq} .= "$id,";
}

while ( my ($seq, $id) = each %seqs) {
    print ">$id\n$seq\n";
}

then you can run it as:

perl removeDuplicates.pl < original.fasta > unique.fasta

[natasha.sernova]

THANK YOU! I was always very much afraid of such complicated data stuctures...

That's great!

Natasha

[JC]

Your welcome, I'm glad to help.