you can give this perl script a try:

#!/usr/bin/env perl

# par1 is seq1 file, par2 is seq2 file

use strict;
use warnings;

my %seq2;
my $id;

open (SEQ2, "< $ARGV[1]");
#read in file 2 , store in hash
while (<SEQ2>) {
        next if ($_ =~ /^\n/);
        if ($_ =~/^>/){
                chomp;
        $id = $_;
                $id =~ s/>//;
        }
        else {
                chomp;
                $seq2{$id} = $_;
        }
}
close SEQ2;

open (SEQ1, "< $ARGV[0]");
while (<SEQ1>) {
        next if ($_ =~ /^\n/);
        if ($_ =~/^>/){
                print $_;
        }
        else {
                chomp;
                my $seq1 = $_;
                my ($repl) = ($seq1 =~ /(\d.+)$/);
                $seq1 =~ s/$repl/\*$seq2{$repl}\*/;
                print $seq1 . "\n";
        }
}
exit;

run it as follows:

perl script.pl seq1File seq2File > outFile

it will print output to outFile. It also assume all the parts to be replaced start with a number.

If the sequence is not too long, you might use grep to find the sequence and save to new file by using >. I'm not quite sure what's your question. Just guessing...