Extracting nucleotide sequence from a non-standardly formatted text file - regex

Question

I have been given some DNA sequences by collaborators in a word document that I'd like to convert into a series of fasta sequences in one file.
I've made it into a text file and I figured that using regular expressions to extract the gene name and the sequence:
use warnings;
use strict;
die "usage: make_fasta.pl <sequence file>" unless (#ARGV == 1);
my $seq_filename = shift;
my $fasta_db_name = $seq_filename . "_db.fa";
open(my $seq_file, '<', $seq_filename)
or die "can't open file $seq_filename, $!";
open(my $fasta_file, '>', $fasta_db_name)
or die "can't open file $fasta_db_name, $!";
while (my $line = <$seq_file>) {
chomp $line;
if ($line =~ /^[ATCG]+$/) { # if the line is entirely DNA seqence
print $fasta_file "$line\n";
} elsif ($line =~ /Full-length (\w+) cDNA/) { # if the line has gene info
print $fasta_file ">$1\n";
} else {
next;
}
}
But that just gave me the name of the first gene. Clearly I've done something wrong with the DNA regular expression but I can't for the life of me work it out. To my eyes it's exactly the same as other suggested DNA tests I've found on this site and others.
The file I'm trying to parse is configured like so:
Collaborators name
title of gene set
Full-length clock cDNA coding sequence
ATGGTAGGATGTGTAATGCGTACGTGATCGT
Full-length per cDNA coding sequence
ATGCTAGCTACGTACGTAGCTACGTAGTACG
I want the output to be a fasta file so:
>clock
ATGGTAGGATGTGTAATGCGTACGTGATCGT
>per
ATGCTAGCTACGTACGTAGCTACGTAGTACG
The first few lines of the actual input file are:
Dr Lin Zhang (Leicester University 10/2012)
Canonical clock genes
Full-length per cDNA coding seq (3693bp)
ATGGACACAGGAACACCCCATGAAGATGTGCCCTCAGAGGACCACACCTTGGAAGAAGGGGACAGCAAGAACCCCTCGTGCCAGCAAGAGTCAGCCTACGGCTCCCTCGAGTCATCCTCCAATGGACAGTCTCAGAAAAGTTTCGGAGGAAGTGGAAGCAAAAGCTTAAATAGTGGTTCGAGTCACAGCAGCGGCTTTGGGGACCAAAATGATTTCAAGGGTATCCATCTTCACGAAGCGAAACACATAGCGTTGAAGAAGAAGAAAACTGGGAAAGGAGGTGAAAAGGTAGCAGAAATCCCCTTTCAAACTGCCTCTGAGGCAGAACTGTCCTCCAAAGGAAACGAAACAGAAAAGGAGAAAGAAACAAGCCTCGAGGAGTCTCCTGCTGCAAAAGAGGAAGCAATTATCGAAAAGGAGTCTCGTTACATCCACCCGAGGAACT

Answer 1

Kind of hard to answer this question without seeing part of the actual input file.
There is a mis-match between your example input and your REGEX:
# looking for verbatim('Full-length') then <space> then one WORD_WITH_ALPHNUMERICS then <space> and then verbatim 'cDNA'
$line =~ /Full-length (\w+) cDNA/;
Your example input line has 'Full length' without a dash, multiple words for the gene name not just one and no 'cDNA' at the end.
If your input line has 'Full-length gene name with multiple words cDNA', your REGEX can be:
$line=~/Full-length\s+(.*?)\s+cDNA/;

Answer 2

The problem is apparently with your input data. I modified the code you posted to produce the following program:
#!/usr/bin/env perl
use warnings;
use strict;
while (my $line = <DATA>) {
chomp $line;
if ($line =~ /^[ATCG]+$/) { # if the line is entirely DNA seqence
print "$line\n";
} elsif ($line =~ /Full-length (\w+) cDNA/) { # if the line has gene info
print ">$1\n";
}
}
__DATA__
Collaborators name
title of gene set
Full-length clock cDNA coding sequence
ATGGTAGGATGTGTAATGCGTACGTGATCGT
Full-length per cDNA coding sequence
ATGCTAGCTACGTACGTAGCTACGTAGTACG
and it produces the output you specified:
~$ src/tmp/cdna
>clock
ATGGTAGGATGTGTAATGCGTACGTGATCGT
>per
ATGCTAGCTACGTACGTAGCTACGTAGTACG
My modifications were only to make it self-contained and did not change any of the flow control or logic, aside from removing the useless else { next } clause.
Can you find and post a few lines of actual data which fails for you, since the dummy data provided seems to work correctly?