[Bioperl-l] cigar string in GenericHSP

Jason Stajich jason at cgt.mc.duke.edu
Wed Mar 12 21:29:22 EST 2003


On Wed, 12 Mar 2003, Bob Freeman wrote:

> At 11:12 AM +0100 3/12/03, Marc Logghe wrote:
> >Hi,
> >I was only wondering, if you really would like to reproduce the original
> >blast alignment, is the original homology string included in the cigar
> >string ? If I recall well, the homology sequence is not included in the aln
> >object and if you want to  add it explicitely, bioperl complains about the
> >bad sequence.
>
> true; the homology sequence is not included. also, trying to get the
> 'match' sequence from $hsp->seq_str('match') doesn't work.
>
Steve and my preferences for how to setup the methods might be out of
whack - try $hsp->homology_string() at least for the GenericHSP, it should
work.

> >Also, I was missing a writer method for blast alignments. I have written my
> >own , but did not yet try to plug it in Bio::SearchIO::blast. If someone
> >else would like to give it a try, I'll include the code I currently have
> >($LINE_LEN is global and should be set to 60):
>
> what about the reportwriter methods? am i missing something here?
>
I think Marc had not found these yet.

> on a separate note, it appears from code that I'm writing that using
> Bio::Tools::Run::StandAloneBlast to run local blast generates
> GenericHit and GenericHSP objects, not the Blast specific ones. E.g.
> print $hit->raw_hit_data throws an error. This method is present in
> Bio::Search::Hit::blast but not Bio::Search::GenericHit. Am I doing
> something wrong?
>

There is a whole history here in the short life of SearchIO - SteveC and I
went about solving this in slightly different ways - hence psiblast and
blast parser - they are being rolled back into one but it means/meant that
there were  2 separate implementations which produced BlastXXX and
GenericXXX objects - see the FAQ and SearchIO HOWTO for more information.
Steve added a whole bunch of methods in his BlastXXX objects which I did
not.

What in raw_hit_data do you want to get back - it is all stored as fields
in either the Hit or HSP object in the GenericXXX implementation.

-j

> Bob >
>

--
Jason Stajich
Duke University
jason at cgt.mc.duke.edu


More information about the Bioperl-l mailing list