You can count the number of occurrences of a particular character in a string with tr:

use 5.010;                                                            
+          

my $str1 = 'MALSSTAATTSSKLKLSNPPSLSHTFTASASASVSNSTSFR';

my $a_count = $str1 =~ tr/A/A/; # 6
[download]

length will give you the length of the entire string.

Now to actually pull out the uppercase sequence from your sample input, are you reading lines from a file? Something like this would probably work:

#!/usr/bin/env perl

use 5.010;

for (<>) {
    if (/^>/) {
        # Header
    } elsif (/^[A-Z]+$/) {
        # Protein
        my $a = tr/A/A/;
        say "A: $a, length: " . length;
    }
}
[download]

Then simply run it with script.pl < protein.txt. Modify the say ... line to taste, or more likely, replace it with the rest of your logic. You can also choose to parse the header if needed, in the # Header section.

You could of course modify this to actually open the file in your script with open instead, if that is more desirable:

open my $fh, '<', $filename
    or die "Couldn't open $filename: $!";

for (<$fh>) {
[download]