Re^4: Comparing a large set of DNA sequences

[roboticus]

BrowserUk:

Thanks!

Yes, this method breaks down pretty quickly when the criteria are relaxed much. But I found a nifty variation that's a little bit more flexible (and still very limited): Rather than replace the character with an asterisk, we can remove the character entirely. That way, not only will it match a single wildcard position, but would allow it to find some two-character differences (such as a character deleted or inserted in two different locations). I need to find a simple way to group the replicated matches before I'm happy with it.

I'm still playing around with it, but I'm at work and need to get stuff done. So I'm posting what I have so far, just in case I never get back around to it:

$ cat 937249.pl
#!/usr/bin/perl
use strict;
use warnings;
my %H;
open my $FH, '<', 'DNA_strings.dat' or die $!;
while (<$FH>) {
    next if /^\s*(#.*)?$/;
    s/\s+$//;
    for my $i (0 .. length($_)-1) {
        my $k = $_;
        substr($k,$i,1) = '';
        $H{$k}{$_}++;
    }
}

for my $k (sort keys %H) {
    if (keys %{$H{$k}} > 1) {
        print "$k\t", join("\n\t\t", keys %{$H{$k}}), "\n";
    }
}

$ cat DNA_strings.dat
# Add a random character to GTTAACCGGA in various positions
GTxTAACCGGA
GTTAAyCCGGA
GTTAACCGzGA

# Delete a random character from GAGGGTGATC in various positions
GAGGGTGAT
GAGGGTGTC
GAGGTGATC
AGGGTGATC

GCAATTTGTC
GCAAATTGTC
GCAATTGGTC
GTTTATAAGT
TGGACAAGCT
TCAGCGGATC
CTACATAACT
TTACTTCAGG
CGGACCTTGG
TGCGTGTGAC

$ perl 937249.pl
AGGGTGAT        GAGGGTGAT
                AGGGTGATC
AGGGTGTC        GAGGGTGTC
                AGGGTGATC
AGGTGATC        AGGGTGATC
                GAGGTGATC
GAGGGTGT        GAGGGTGAT
                GAGGGTGTC
GAGGTGAT        GAGGGTGAT
                GAGGTGATC
GAGGTGTC        GAGGGTGTC
                GAGGTGATC
GCAATTGTC       GCAATTTGTC
                GCAATTGGTC
                GCAAATTGTC
GGGTGATC        AGGGTGATC
                GAGGTGATC
GTTAACCGGA      GTTAAyCCGGA
                GTxTAACCGGA
                GTTAACCGzGA

I'd like to eliminate the duplicate groups, probably by building another hash from the original results. But I haven't figured out a nice way to do so yet.

...roboticus

When your only tool is a hammer, all problems look like your thumb.

Update: Minor formatting & text update.