Re^2: Lower-casing Substrings and Iterating Two Files together

[neversaint]

Dear BrowserUK,
Your magical approach is at least 6980x faster than mine!

My approach:

191.52user 153.47system 5:49.44elapsed 98%CPU 

BrowserUk Perl's approach (modified to handle FASTA):

open SEQ,  '<', $ARGV[0] or die $!; #plain
open MASK, '<', $ARGV[1] or die $!; #hardmask

while ( my $seq = <SEQ> ) {    ## Read a sequence 
    my $mask = <MASK>;         ## And the corresponding mask

    if ( $mask =~ /^>/ ) {    
        print "$seq";
    }
    else {
        $mask =~ tr[N][ ];         ## Ns => spaces
        print $seq | $mask;        ## bitwise-OR them and print the re
+sult;
        
    }
    
}
close SEQ;
close MASK;

Takes only this much time:

0.45user 0.07system 0:05.59elapsed 9%CPU 

This is tested on 12MB dataset, and the breakdown of sequence length is this:

#Seq_name    #Seq_len
2-micron    6318
MT    85779
I    230208
VI    270148
III    316617
IX    439885
VIII    562643
V    576869
XI    666454
X    745745
XIV    784333
II    813178
XIII    924429
XVI    948062
XII    1078175
VII    1090946
XV    1091289
IV    1531918

Update: These test datasets (full and hardmasked) can be downloaded here.