[Bioperl-l] Making BioPerl Faster

Chris Fields cjfields at uiuc.edu
Mon Jul 24 16:04:40 UTC 2006


Give it a look, sure!  Not sure if this the only problem though when it
comes to speed; I think it's more complicated than that.  I think that (at
least on WinXP) the Perl version used is also partially to blame.  It's
possible that something modified between v 5.6 and 5.8 slowed everything
down considerably.  I always wondered if it had something to do with Unicode
support in perl 5.8 ...

There is a report on Bugzilla about a dramatic slowdown on sequence parsing
between v. 1.4 and v. 1.5 (including the latest, v 1.5.1)

http://bugzilla.open-bio.org/show_bug.cgi?id=1875

This is unresolved at this time but may be unrelated to the possible perl
versioning issue above.

I've a feeling you may find regexes and redundant methods calls also add
quite a bit of overhead.  I've seen several places where accessors are
called over and over w/o assigning to a local variable.  Or places where a
tr/// would work much faster than a s///.  There was an instance of the
latter in SeqIO which sped up parsing about 2-3x faster on WinXP.

If you want to look at the impact of object instantiation on speed, check
out Bio::SearchIO (parsing of BLAST/FASTA/HMMER reports).  Lots of method
calls, object creation, etc.

Chris

> -----Original Message-----
> From: bioperl-l-bounces at lists.open-bio.org [mailto:bioperl-l-
> bounces at lists.open-bio.org] On Behalf Of Al Ramsey
> Sent: Monday, July 24, 2006 9:24 AM
> To: bioperl-l at lists.open-bio.org
> Subject: [Bioperl-l] Making BioPerl Faster
> 
> I'm interested into following up with a suggestion from the bioperl.org
> site about making it faster
> (http://www.bioperl.org/wiki/Why_BioPerl_is_slow).  In particular, I
> wanted to look a little more into how the object instantiations might be
> more efficient.  Is anyone else looking into this actively now?  I want
> to ask if anyone had any additional insights that weren't previously
> published before I started.
> 
> Thank you,
> Al Ramsey
> 
> 
> --
> Alvin Ramsey, PhD.
> 
> Vecna Technologies, Inc.
> 5205 Leesburg Pike
> Falls Church, VA  22041
> aramsey at vecna.com
> t: 703.998.5333
> f: 703.998.5816
> 
> 
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l




More information about the Bioperl-l mailing list