[Bioperl-l] SearchIO speed up
cjfields at uiuc.edu
Fri Aug 18 11:39:18 EDT 2006
> On 8/18/06 9:52 AM, "Hilmar Lapp" <hlapp at gmx.net> wrote:
> > On Aug 18, 2006, at 3:00 AM, Sendu Bala wrote:
> > I.e., I don't think Bioperl, or any other toolkit that as its main
> > benefits offers a consistent API, object model, tried-and-tested
> > parsers, etc, should try to be the natural choice for a use case when
> > you don't need any of these benefits.
> I can't agree more here. There are many cases with biological data and
> bioinformatics where bioperl is just not the right answer, and that is OK!
True. I agree that, if speed is an issue, BioPerl may not be the best
solution. But I think it is feasible to make SearchIO faster. Just not the
way it is currently implemented.
I think that our focus also needs to include SeqIO. Some significant
sequence parsing slowdowns were noted between bioperl 1.4 and bioperl 1.5.1,
probably tied to the addition of several new classes. There may not be a
way around this except SwissKnife-like 'lazy parsing.'
Hilmar's wiki idea for Bioperl 2.0 (Warp9?) is already in place in a way:
there is a 'Beyond' section for Bioperl Releases, so it wouldn't be too hard
to set up a link to a new page for BioPerl 2.0 ideas.
More information about the Bioperl-l