Find all repeated 4-mers in a DNA Sequence - Perl

Question

Hello,

I try to write a program that reads in a FASTA-formatted file containing multiple DNA sequences, identifies all repeated 4-mers (i.e., all 4-mers that occur more than once) in a sequence, and prints out the repeated 4-mer and the header of the sequence in which it was found. A k-mer is simply a sequence of k nucleotides (e.g., “aaca”, “gacg”, and “tttt” are 4-mers).

Here's my code:

use strict;
use warnings;

my $count = -1;
my $file = "sequences.fa";
my $seq = '';
my @header = ();
my @sequences = ();
my $line = '';
open (READ, $file) || die "Cannot open $file: $!.\n";

while ($line = <READ>){
    chomp $line;
    if ($line =~ /^>/){
        push @header, $line;
        $count++;
        unless ($seq eq ''){
            push @sequences, $seq;
            $seq = '';
        }
    } else {
        $seq .= $line;
    }
}   push @sequences, $line;

for (my $i = 0; $i <= $#sequences+1; $i++){
    if ($sequences[$i] =~ /(....)(.)*\g{1}+/g){
        print $header[$i], "\n", $&, "\n";
    }
}

I have two requests: First, I don't know how to design my regex pattern to get the desired output. And second, less importantly, I'm sure my code is very inefficient, so if there's a way to shorten it, please tell me.

Thanks in advance!

Here's an example for a FASTA file: (Note that there's an extra line between the sequences, which is NOT the case in original fasta files)

>NC_001422.1 Enterobacteria phage phiX174 sensu lato, complete genome GAGTTTTATCGCTTCCATGACGCAGAAGTTAACACTTTttttttCGGATATTTCTGATGAGTCGAAAAAT CCCTTACTTGAGGATAtatataAATTATGTCTAATATTCAAACTGGCGCCGAGCGTATGCCGCATGACCT

>NC_001501.1 Enterobacteria phage phiX184 sensu lato, complete genome AACGGCTGGTCAGTATTTAAGGTTAGTGCTGAGGTTGACTACATCTGTTTTTAGAGACCCAGACCTTTTA TCTCACTTCTGTTACTCCAGCTTCTTCGGCACCTGTTTTACAGACACCTAAAGCTACATCGTCAACGTTA TATTTTGATAGTTTGACGGTTAATGCTGGTAATGGTgagagagaGGTTTTCTTCATTGCATTCAGATGGA TCAACGCCGCTAATCAGGTTGTTTCTGTTGGTGCTGATATTGCTTTTGATGCCGACCCTAAATTTTTTGC CTGTTTGGTTCGCTTTGAGTCTTCTTCGGTTCCGACTACCCTCCCGACTGCCTATGATGTTTATCCTTTG

>NC_001622.5 Enterobacteria phage phiX199 sensu lato, complete genome TTCGCTGAATCAGGTTATTAAAGAGTTGCCGAGATATTTATGTTGGTTTCATGCGGATTGGTCGTTTAAA TTGGACTTGGTGGCAAGTCTGCCGCTGATAAAGGAAAGGATAATGACCAAATCAAAGAACTCGTGATTAT CTTGCTGCTGCATTTCCTGAGCTTAATGCTTGGGAGCGTGCTGGTGCTGATGCTTCCTCTGCTGGTATGG TTGACGCCGGATTTGAGAATCAAAAATGTGAGAGAGCTTACTAAAATGCAACTGGACAATCAGAAAGAGA GATGCAAAATGAGACTCAAAAAGAGATTGCTGGCATTCAGTCGGCGACTTCACGCCAGAATACGAAAGAC CAGGTATATGCACAAAATGAGATGCTTGCTTATCAACAGAAGGAGTCTACTGCTCGCGTTGCGTCTATTA TGGAAAACACCAATCTTTCCAAGCAACAGCAGGTTTCCGAGATTATGCGCCAAATGCTTACTCAAGCTCA AACGGCTGGTCAGTATTTTACCAATGACCAAATCAAAGAAATGACTCGCAAGGTTAGTGCTGAGGTTGAC TTAGATGAGTGTTCATCAGCAAACGCAGAATCAGCGGTATGGCTCTTCTCATATTGGCGCTACTGCAAAG

Answer 1

I'd probably tackle your problem rather more like this:

#!/usr/bin/env perl

use strict;
use warnings;

use Data::Dumper;

#set paragraph mode. Iterate on blank lines. 
local $/ = ''; 

#read from STDIN or a file specified on command line, 
#e.g. cat filename_here | myscript.pl
#or myscript.pl filename_here
while ( <> ) {
   #capture the header line, and then remove it from our data block
   my ($header) = m/\>(.*)/;
   s/>.*$//;

   #remove linefeeds and whitespace. 
   s/\s*\n\s*//g;
   #use lookahead pattern, so the data isn't 'consumed' by the regex. 
   my @sequences = m/(?=([atcg]{4}))/gi;

   #increment a count for each sequence found. 
   my %count_of;
   $count_of{$_}++ for @sequences;

   #print output. (Modify according to specific needs. 
   print $header,"\n";

   print "Found sequences:\n";
   print Dumper \@sequences;
   print "Count:\n";
   print Dumper \%count_of;

   #note - ordered, but includes duplicates. 
   #you could just use keys  %count_of, but that would be unordered. 
   foreach my $sequence ( grep { $count_of{$_} > 1 } @sequences ) {
      print $sequence, " => ", $count_of{$sequence},"\n";
   }
   print "\n";
}

We iterate record by record, capture and remove the 'header' line, and then splice together the rest. Then capture each (overlapping) sequence of 4, and count them.

This, for your sample data (first stanza for brevity):

NC_001422.1 Enterobacteria phage phiX174 sensu lato, complete genome 
Found sequences:
    GAGT => 2
    AGTT => 2
    TTAT => 2
    CATG => 2
    ATGA => 3
    TGAC => 2
    CGCA => 2
    AGTT => 2
    ACTT => 2
    tttt => 3
    tttt => 3
    tttt => 3
    GGAT => 2
    GATA => 2
    ATAT => 2
    TATT => 2
    ATGA => 3
    TGAG => 2
    GAGT => 2
    AAAA => 2
    AAAA => 2
    ACTT => 2
    TGAG => 2
    GGAT => 2
    GATA => 2
    tata => 2
    tata => 2
    TTAT => 2
    TATG => 2
    ATAT => 2
    TATT => 2
    GCCG => 2
    TATG => 2
    GCCG => 2
    CGCA => 2
    CATG => 2
    ATGA => 3
    TGAC => 2

Note - because it's based on the original sequences, it's based on ordering within the data, and you'll see TGAC there twice because ... it's in there twice.

However you could instead:

   foreach my $sequence ( sort { $count_of{$b} <=> $count_of{$a} }
                          grep { $count_of{$_} > 1 } 
                                 keys %count_of ) {
      print $sequence, " => ", $count_of{$sequence},"\n";
   }
   print "\n";

Which will discard any with less than 2 matches, and order by frequency.