My toughts, too. Regex's seem like an overly complicated solution to me; but, on the other hand, it's a good problem to learn some pretty empowering Regex strategies. So "Your mileage may vary."

Here is a quick sript that uses substr() and index() to find the search-for pattern and deletes everything up to the pattern plus that pattern. It also only searches up to the specified number of characters.

#!/usr/bin/perl

use strict;
use warnings;

my @strings   = ('ATGGCTAGTTTCAGT',
                 'GTTTATGGGTACGTGT',
                 'ATTGATACGGGTATTTAGGCTG',
                 'ATTAGGAAGTAGGACCCTAGG'
                 );
my $search       = 'TTT';
my $len = length($search);

my @newStrings = map {my $indx = index(substr($_,0,8),$search);
                      my $temp;
                      if($indx == -1){
                            $temp = $_;
                      }else{
                            $temp = substr($_,
                                          ($indx+$len),(length($_)-$in
+dx-$len));
                      }
                      } @strings;

foreach my $indx1 (0..(scalar(@strings)-1)){
   print "$strings[$indx1] ---> $newStrings[$indx1]\n";
}

exit(0);
[download]

This script produces the following output:

ATGGCTAGTTTCAGT ---> ATGGCTAGTTTCAGT
GTTTATGGGTACGTGT ---> ATGGGTACGTGT
ATTGATACGGGTATTTAGGCTG ---> ATTGATACGGGTATTTAGGCTG
ATTAGGAAGTAGGACCCTAGG ---> ATTAGGAAGTAGGACCCTAGG
[download]

If I understand the OP's inquiry, I think this does what is wanted.

The only thing I'm not clear on is what the OP expects to do if there are (1) multiple occurances, within the seek-or number of characters, of the search-for pattern in the string? and (2) what the result should be if there is a run of the search-for pattern (e.g., 'TTTTTTTTT' which has three intances of the search-for pattern in a run)?