[Bioperl-l] frac_aligned_query returning results >1.

Steve Chervitz sac at bioperl.org
Fri Mar 2 18:00:28 EST 2007


Glad you fixed the problem, Sendu.

I thought this might have been due to a problem in HSPI::matches() since it
was reporting (1507,1507) as an invalid range within (1444,1507), when it
should be valid (the last position). So it looked like an edge condition
bug, but I didn't confirm. So there still could be a lingering problem in
the matches() function, or in the way the matches string is parsed from the
report.

Speaking of which, HSPI::matches() is quite BLAST-specific. It's even format
specific, since it won't work if you are parsing in tabular blast reports as
they lack any string of match symbols. I thought about moving the matches
implementation in HSPI into BlastHSP.pm, but that module appears to not be
used anymore. Not sure the way to go here.

Steve

On 3/2/07, Thiago Venancio <thiago.venancio at gmail.com> wrote:
>
> Hi Sendu,
>
> Great to know you fixed the problem.
> I have updated the SearchUtils and seems to be correct now.
>
> Best!
>
> Thiago
>
>
> On 3/2/07, Sendu Bala <bix at sendu.me.uk> wrote:
> >
> > Thiago Venancio wrote:
> > > Hi Sendu and Chris,
> > >
> > > Thanks for the help.
> > > As I mentioned, I have updated my SearchUtils file from:
> > >
> >
> http://code.open-bio.org/cgi-bin/viewcvs/viewcvs.cgi/bioperl-live/Bio/Search/SearchUtils.pm
> > > <
> >
> http://code.open-bio.org/cgi-bin/viewcvs/viewcvs.cgi/bioperl-live/Bio/Search/SearchUtils.pm
> > >
> > >
> > > I am also using the lates BioPerl version, installed from CPAN.
> > >
> > > Please find a buggy blast report attached.
> > > In this case, the frac_aligned_query() outputs "1.04", but I have
> others
> > > with " 1.57" for example.
> > >
> > > Just for a quantitative aspect, I got ">1" values in only 61 / 53,377.
> >
> > Many thanks for that.
> >
> > I've committed another fix for SearchUtils so please get revision 1.23
> > and try again. Hopefully all 61 will no longer be >1, but if any are
> > please send me sample blast files again.
> >
> > For anyone interested, the bug was due to a completely unbelievable
> > oversight on my part in the contig merging algorithm: I forgot to deal
> > with contigs that were fully contained by others. Wow!
> >
>
>
>
> --
> "The way to get started is to quit talking and begin doing."
>       Walt Disney
>
> ========================
> Thiago Motta Venancio, MSc
> PhD student in Bioinformatics
> University of Sao Paulo
> ========================
> _______________________________________________
> 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