basic regex and string manipulation for DNA analysis using perl - perl

Question

I am new to perl and would like to do what I think is some basic string manipulation to DNA sequences stored in an rtf file.
Essentially, my file reads (file is in FASTA format):
>LM1
AAGTCTGACGGAGCAACGCCGCGTGTATGAAGAAGGTTTTCGGATCGTAA
AGTACTGTCCGTTAGAGAAGAACAAGGATAAGAGTAACTGCTTGTCCCTT
GACGGTATCTAACCAGAAAGCCACGGCTAACTACGTGCCAGCAGCCGCGG
TAATACGTAGGTGGCAAGCGTTGTCCGGATTTATTGGGCGTAAAGCGCGC
GCAGGCGGTCTTTTAAGTCTGATGTGAAAGCCCCCGGCTTAACCGGGGAG
GGTCATTGGAAACTGGAAGACTGGAGTGCAGAAGAGGAGAGTGGAATTCC
ACGTGTAGCGGTGAAATGCGTAGATATGTGGAGGAACACCAGTGGCGAAG
GCGACTCTCTGGTCTGTAACTGACGCTGAGGCGCGAAAGCGTGGGGAGCA
AACAGGATTAGATACCCTGGTAGTCCACGCCGT
What I would like to do is read into my file and print the header (header is >LM1) then match the following DNA sequence GTGCCAGCAGCCGC and then print the preceding DNA sequence.
So my output would look like this:
>LM1
AAGTCTGACGGAGCAACGCCGCGTGTATGAAGAAGGTTTTCGGATCGTAA
AGTACTGTCCGTTAGAGAAGAACAAGGATAAGAGTAACTGCTTGTCCCTT
GACGGTATCTAACCAGAAAGCCACGGCTAACTAC
I have written the following program:
#!/usr/bin/perl
use strict; use warnings;
open(FASTA, "<seq_V3_V6_130227.rtf") or die "The file could not be found.\n";
while(<FASTA>) {
chomp($_);
if ($_ =~ m/^>/ ) {
my $header = $_;
print "$header\n";
}
my $dna = <FASTA>;
if ($dna =~ /(.*?)GTGCCAGCAGCCGC/) {
print "$dna";
}
}
close(FASTA);
The problem is that my program reads the file line by line and the output I am receiving is the following:
>LM1
GACGGTATCTAACCAGAAAGCCACGGCTAACTAC
Basically I don't know how to assign the entire DNA sequence to my $dna variable and ultimately don't know how to avoid reading the DNA sequence line by line. Also I am getting this warning:
Use of uninitialized value $dna in pattern match (m//) at stacked.pl line 14, line 1113.
If anyone could give me some help with writing better code or point me in the correct direction it would be much appreciated.

Answer 1

Using the pos function:
use strict;
use warnings;
my $dna = "";
my $seq = "GTGCCAGCAGCCGC";
while (<DATA>) {
if (/^>/) {
print;
} else {
if (/^[AGCT]/) {
$dna .= $_;
}
}
}
if ($dna =~ /$seq/g) {
print substr($dna, 0, pos($dna) - length($seq)), "\n";
}
__DATA__
>LM1
AAGTCTGACGGAGCAACGCCGCGTGTATGAAGAAGGTTTTCGGATCGTAA
AGTACTGTCCGTTAGAGAAGAACAAGGATAAGAGTAACTGCTTGTCCCTT
GACGGTATCTAACCAGAAAGCCACGGCTAACTACGTGCCAGCAGCCGCGG
TAATACGTAGGTGGCAAGCGTTGTCCGGATTTATTGGGCGTAAAGCGCGC
GCAGGCGGTCTTTTAAGTCTGATGTGAAAGCCCCCGGCTTAACCGGGGAG
GGTCATTGGAAACTGGAAGACTGGAGTGCAGAAGAGGAGAGTGGAATTCC
ACGTGTAGCGGTGAAATGCGTAGATATGTGGAGGAACACCAGTGGCGAAG
GCGACTCTCTGGTCTGTAACTGACGCTGAGGCGCGAAAGCGTGGGGAGCA
AACAGGATTAGATACCCTGGTAGTCCACGCCGT
You can process a file with multiple entries like so:
while (<DATA>) {
if (/^>/) {
if ($dna =~ /$seq/g) {
print substr($dna, 0, pos($dna) - length($seq)), "\n";
$dna = "";
}
print;
} elsif (/^[AGCT]/) {
$dna .= $_;
}
}
if ($dna && $dna =~ /$seq/g) {
print substr($dna, 0, pos($dna) - length($seq)), "\n";
}

Answer 2

Your while statement reads until the end of file. That means at every loop iteration, $_ is the next line in <FASTA>. So $dna = <FASTA> isn't doing what you think it is. It is reading more than you probably want it to.
while(<FASTA>) { #Reads a line here
chomp($_);
if ($_ =~ m/^>/ ) {
my $header = $_;
print "$header\n";
}
$dna = <FASTA> # reads another line here - Causes skips over every other line
}
Now, you need to read the sequence into your $dna. You can update your while loop with an else statement. So if its a head line, print it, else, we add it to $dna.
while(<FASTA>) {
chomp($_);
if ($_ =~ m/^>/ ) {
# It is a header line, so print it
my $header = $_;
print "$header\n";
} else {
# if it is not a header line, add to your dna sequence.
$dna .= $_;
}
}
After the loop, you can do your regex.
Note: This solution assumes there is only 1 sequence in the fasta file. If you have more than one, your $dna variable will have all the sequences as one.
Edit: Adding simple a way to handle multiple sequences
my $dna = "";
while(<FASTA>) {
chomp($_);
if ($_ =~ m/^>/ ) {
# Does $dna match the regex?
if ($dna =~ /(.*?)GTGCCAGCAGCCGC/) {
print "$1\n";
}
# Reset the sequence
$dna = "";
# It is a header line, so print it
my $header = $_;
print "$header\n";
} else {
# if it is not a header line, add to your dna sequence.
$dna .= $_;
}
}
# Check the last sequence
if ($dna =~ /(.*?)GTGCCAGCAGCCGC/) {
print "$1\n";
}

Answer 3

I came up with a solution using BioSeqIO (and the trunc method from BioSeq from the BioPerl distribution. I also used index to find the subsequence rather than using a regular expression.
This solution does not print out the id, (line begins with >), if the subsequence was not found or if the subsequence begins at the first postion, (and thus no preceding characters).
#!/usr/bin/perl
use strict;
use warnings;
use Bio::SeqIO;
my $in = Bio::SeqIO->new( -file => "fasta_junk.fasta" ,
-format => 'fasta');
my $out = Bio::SeqIO->new( -file => '>test.dat',
-format => 'fasta');
my $lookup = 'GTGCCAGCAGCCGC';
while ( my $seq = $in->next_seq() ) {
my $pos = index $seq->seq, $lookup;
# if $pos != -1, ($lookup not found),
# or $pos != 0, (found $lookup at first position, thus
# no preceding characters).
if ($pos > 0) {
my $trunc = $seq->trunc(1,$pos);
$out->write_seq($trunc);
}
}
__END__
*** fasta_junk.fasta
>LM1
AAGTCTGACGGAGCAACGCCGCGTGTATGAAGAAGGTTTTCGGATCGTAA
AGTACTGTCCGTTAGAGAAGAACAAGGATAAGAGTAACTGCTTGTCCCTT
GACGGTATCTAACCAGAAAGCCACGGCTAACTACGTGCCAGCAGCCGCGG
TAATACGTAGGTGGCAAGCGTTGTCCGGATTTATTGGGCGTAAAGCGCGC
GCAGGCGGTCTTTTAAGTCTGATGTGAAAGCCCCCGGCTTAACCGGGGAG
GGTCATTGGAAACTGGAAGACTGGAGTGCAGAAGAGGAGAGTGGAATTCC
ACGTGTAGCGGTGAAATGCGTAGATATGTGGAGGAACACCAGTGGCGAAG
GCGACTCTCTGGTCTGTAACTGACGCTGAGGCGCGAAAGCGTGGGGAGCA
AACAGGATTAGATACCCTGGTAGTCCACGCCGT
*** contents of test.dat
>LM1
AAGTCTGACGGAGCAACGCCGCGTGTATGAAGAAGGTTTTCGGATCGTAAAGTACTGTCC
GTTAGAGAAGAACAAGGATAAGAGTAACTGCTTGTCCCTTGACGGTATCTAACCAGAAAG
CCACGGCTAACTAC

Answer 4

read the whole file into memory then look for the regexp
while(<FASTA>) {
chomp($_);
if ($_ =~ m/^>/ ) {
my $header = $_;
print "$header\n";
} else {
$dna .= $_;
}
}
if ($dna =~ /(.*?)GTGCCAGCAGCCGC/) {
print $1;
}