Find the postion of SNP in the gen list

Question

I have SNP data and gen list data. I am looking for the position of SNP cotain in the gen list data when I compare with gen list. For example:

1. The SNP data :

   Pos_start pos_end 
   14185     14185      
   ....      .....   
   

2. The gen list data:

   5"side(pos_start)  3"sile(pos_end)
     1                  1527      
     1920               1777 
     ....               ..... 
   

3. the result: in the position 14185 of SNP contain at the 16185 position of the gen list.

Below is my code but it has some problem in sort the number.

   #!/usr/bin/perl -w

   open(POS1,"<posi1.txt"); (I collect two data and save with posi1.txt) 
    @posi1=<POS1>;
   open(list,">list.txt");
   @list1=@posi1;
   @list2= sort num_last (@list1);
   $list2 = join( '', @list2);

   print $list2;
   print list $list2."\n\n";
   close(list);
  sub num_last {
my ($num_a, $num_b);
$num_a=$a=~ /＾[0-9]/;
$num_b=$b=~ /＾[0-9]/;
if ($num_a && $num_b){
    return $a<=>$b;
} elsif ($num_a){
    return 1;
} elsif ($num_b){
    return -1;
} else {
    return $a cmp $b;
}
      }

I would appreciate if you could give some pointers.

Answer 1

First of all, your sort sub does not operate on values you pass. It should be something like

sub num_last {
    my ($num_a, $num_b);
    my ($a,$b) = @_;
    ....
}

Than, you are really getting only first digit in a string if the string starts from digit. It's better add skipping all leading whitespaces, just in case.

($num_a) = $a =~ /^\s*(\d+)/;
($num_b) = $b =~ /^\s*(\d+)/;

\d+ is equivalent to [0-9]+, but two chars shorter :). Braces force list context so, $num_a and $num_b receives content of first matched group: (\d+).

Than, you don't need <=> opertor, as $num_a and $num_b should be strings, so you can simplify your condition to:

if (!$num_a)
    return -1;
if (!$num_b)   
    return 1;
return $a cmp $b;

Not sure, but it might be as simple as return $a cmp $b, but I'm not sure if empty var is stringwise lesser than non-empty string and no perl at fingertips. So, final num_last function:

sub num_last{
    my ($num_a, $num_b);
    my ($a,$b) = @_;

    ($num_a) = $a =~ /^\s*(\d+)/;
    ($num_b) = $b =~ /^\s*(\d+)/;

    if (!$num_a)
        return -1;
    if (!$num_b)   
        return 1;
    return $a cmp $b;
}

If you need reverse sort, just replace my ($a,$b) = @_; with my ($b,$a) = @_;

And, I've written it without any compiler help, so there might be some minor errors in it.