Beefy Boxes and Bandwidth Generously Provided by pair Networks
laziness, impatience, and hubris
 
PerlMonks  

Re^3: Mismatch Positions of Ambiguous String

by hv (Parson)
on Apr 27, 2006 at 10:16 UTC ( #545996=note: print w/ replies, xml ) Need Help??


in reply to Re^2: Mismatch Positions of Ambiguous String
in thread Mismatch Positions of Ambiguous String

But doesn't work. Where did I go wrong?

It isn't as simple as that: my original approach relied on treating each ambiguous [ACGT] in the source string as a regexp, but to do that it needs to know a) which is the fixed string and which the regexp, and b) that they are not both regexps.

The simplest extension is as below, but this suffers further on speed - it will probably be too slow if you're dealing with strings a few thousand base pairs long:

sub mismatches { my($source, $target) = @_; my @sparts = ($source =~ /(\[.*?\]|.)/g); my @tparts = ($target =~ /(\[.*?\]|.)/g); scalar grep { my($s, $t) = ($sparts[$_], $tparts[$_]); $s !~ /\[/ ? ($s !~ /$t/) : $t !~ /\[/ ? ($t !~ /$s/) : !intersect($s, $t) } 0 .. $#sparts; } sub intersect { my($s, $t) = @_; my %seen = map +($_ => 1), $s =~ /[^\[\]]/g; scalar grep $seen{$_}, $t =~ /[^\[\]]/g; }

This says: if source is not ambiguous, treat the corresponding fragment of the target as a regexp; else if the target is not ambiguous, treat the source fragment as a regexp; else check a full intersection of the two.

If your strings only include ACGT, a more efficient approach would be to transform each string into a bit vector that sets a bit for each base that may be present:

my %bits = ('A' => 1, 'C' => 2, 'G' => 4, 'T' => 8); my $source1 = bitwise('[TCG]GGGG[AT]'); my $target1 = bitwise('AGGGG[CT]'); print mismatches($source, $target1), "\n"; sub mismatches { my($source, $target) = @_; ($source & $target) =~ tr/\0//; } sub bitwise { my $string = shift; join '', map { my $char = 0; $char |= $bits{$_} for /[ACGT]/g; chr($char) } $string =~ /(\[.*?\]|.)/g; }

Once the strings are transformed into this bitwise representation, checking for mismatches is very fast even with long strings.

Hugo


Comment on Re^3: Mismatch Positions of Ambiguous String
Select or Download Code

Log In?
Username:
Password:

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://545996]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others chanting in the Monastery: (12)
As of 2014-08-27 18:41 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    The best computer themed movie is:











    Results (248 votes), past polls