why can't i use dnacmp2? it seems to work just fine with the above code

As I mentioned above, dnacmp() produces a list of mismatches, (Eg. "1:A:C", "2:A:G").

So in the first version of the counting code I posted:

m[\d+:(.:.)] and ++$stats{ $1 } for dnacmp( @seqs[ $i, $j ] );
[download]

I iterate over that list (using for), and m///atch against each individual mismatch in order to count.

But dnacmp2 returns a single string containing all the mismatches, (Eg. "1:A:C 2:A:G"), so the above method would only see and count the first mismatch.

In the latest code, I use:

++$stats{ $1 } while $res =~ m[:(.:.)]g;
[download]

Which uses a while loop and m///g to count all the mismatches in the single string.