If/else within a foreach loop to check strings within an array

[Anonymous Monk]

Hi -- I'm working on writing code to check each string within an array. This is directed at checking that a DNA input contains only A, T, C, and G. I have written it such that you can input your DNA sequence on the command line. When I run this with an input such as "ATGTCAHCGT," I get nothing back where I should get a print statement of "Non-nucleotide at 6 position." What I am missing/ where am I going wrong? Any help much appreciated. Cheers, A

my $dna_input = $ARGV[0]; # turn command line input into a variable

my @dna = split("", $dna_input); #split the variable into an array so 
+as to check each nucleotide

my $count = 0; # keep track of position so as to report where there is
+ a non-nuceloetide character
my $base;
foreach $base (@dna) {
        if($base eq 'A' || 'T' || 'C' || 'G') {
                $count += 1;
                }
        else {
                 print "Non-nucleotide at $count position\n";
                }
}

[choroba]

Very similar to a recent question. See Re: Trying to compare a string....

Update: In your case, regular expression would be the easiest solution:

my $dna = shift; # Get the first command line argument.
if ($dna =~ /[^ACTG]/g) {
    die "Non-nucleotide at " . pos($dna);
}

[Anonymous Monk]

Fantastic, thank you!

[LanX]

you are getting precedence wrong, what you are effectively doing is ($base eq 'A') || 'T' || 'C' || 'G'

There are plenty of alternatives,like:

$base eq 'A' || $base eq 'T' || $base eq 'C' || $base eq 'G'

[BrowserUk]

Whilst for the very short input in your example choroba's solution is perfectly fine, for real sequences which are typically very large, spliting the sequence to an array of scalars each holding a single char and then processing them one at a time is grossly inefficient. Of both time and memory.

Far more efficient for your stated purpose is to process the sequence as a string:

my $dna_input = $ARGV[0];
print "Non-nucleotide '$1' at posn %d\n", $-[0] while $dna_input =~ m[
+([^acgt])]g;

[LanX]

> Whilst for the very short input in your example choroba's solution is perfectly fine,

Read again! No big difference to your approach.

[BrowserUk]

You're correct. My mistake.

[drmrgd]

It is possible to have more than non-standard nucleotide in your sequence? If so, you might consider looking for all the non-matches:

#!/usr/bin/perl

use warnings;
use strict;

my $dna_input = shift;

my @dna = split( '', $dna_input );
my (@index) = grep { $dna[$_] !~ /[ATCG]/ } 0..$#dna;
print "Non-nucleotide at position $_\n" for @index;

[boftx]

getc comes to mind for char-at-a-time if one needs to process a file with potentially very long lines (or no line/record endings at all) instead of taking input directly as a command line arg.