Your regex is using greedy matching, so your first match term is 'Name=ARF1; Synonyms'. You can make it less greedy using '+?'.

however:

This won't fix your problem because your format requires multiple passes per line, and you are only performing one. Perhaps something like this?

@lines =  grep {/^DE|^GN|^ID/} split ("\n", $_);
foreach $lines(@lines) {
    if ($lines =~ /^DE|^GN/ && $lines !~ /Putative uncharacterized pro
+tein/) {
        while ($lines) {
            $lines =~ s/.+?\=(.+?)\;//;
            print lc($1)."\n";
        }
    } elsif    ($lines =~ /^ID/)    {
        print " \n";
    }
}
[download]