use strict;
use warnings;

use Data::Dumper;

my %replace = (
   R => '{A,G}',
   S => '{C,G}',
   K => '{G,T}',
   ' ' => '_',
);

my $s = 'KAG GTR CAG CTG AAG SAG TCA GG';

my @results;

for my $base (keys %replace) {
    $s =~ s/$base/$replace{$base}/g;
}
push @results, $_ while glob $s;
@results = map { s/_/ /g; $_ } @results;
[download]

Genius!! Anything I'd have written would have been considerably longer! Thanks very much.

use strict;

#the substitution possibilities
my @a = ('a');
my @b = ('d','e');
my @c = ('f','g','h');

my $string = pop(@_);  #this is an argument passed to the script on th
+e command line
#my $string = "abc";   #if you want it hardcoded

my $expanded_string = "";
my @strings = ();
foreach my $first (@a){
  foreach my $second (@b){
    foreach my $third (@c){
      $string = "$first$second$third";
      push(@strings, $string); #to store them all
      print "$string\n"; #to print this particular one
    }
  }
}
[download]

Thanks, all the replies are helping me get there but it's still a bit of a nightmare to code. Here's some example inputs etc.

R = A or G
S = C or G
K = G or T
All the other letters stay constant.

my $seq1 = "CAG GTR CAG CTG AAG SAG TCA GG";
my $seq2 = "GAK GTG CAG CTT CAG CAG TCR GG";

The gaps between sets of 3 letters aren't important - it just signifies DNA codons.

So, both seq 1 and seq 2 have 4 possible resulting sequences. I need to store the 4 seqs associated with seq 1 separately e.g. in a different array to those of seq 2.

If you have any ideas about the best way to do this I'd be very grateful! Speed isn't a big consideration, as long as it works!

Thanks for your help.