[Bioperl-l] blasting with multiple query sequences?

Jason Stajich jason@cgt.mc.duke.edu
Fri, 4 Oct 2002 08:53:06 -0400 (EDT)


$report->query_name
$report->query_description

On Fri, 4 Oct 2002, Tobias Thierer wrote:

> Hi,
>
> Bio::Tools::Run::StandAloneBlast::blastall
>
> accepts a reference to an array of Bio::Seq objects.
> If such a reference is passed, all the sequences in
> the array will be blasted against the specified database.
>
> However, although I verbosely browsed the documentation,
> I could not find any hint on how to get the sequence
> that belongs to a specific Subject or HSP. The BLAST
> report gives me all HSPs grouped by subject, but it
> seems that I have no way to actually find out which of
> the sequences that I passed to blastall() actually caused
> the hit.
>
> I'd really appreciate if someone could tell me if and
> how this is possible. I need to blast multiple sequences
> against a database, and get the high scoring matches for
> each query sequence. The blast report, however, groups
> the blast result by subject, where each subject consists
> of a number of HSPs. $hsp->query is only a
> Bio::SeqFeature::Similarity object, not the original
> sequence object. If I pass only one sequence to
> blastall() at a time, the blast executable will be run
> multiple times and thus the database will innecessarily
> be loaded multiple times from the hard disk into the
> memory.
>
> Any help would really be appreciated!
>
> Regards,
>
>   Tobias
>
>
> P.S.: Here is what I am currently doing:
>
>   # ($sequences is a reference to an array of Bio::Seq
>   # objects)
>
>   my $report = $blaster->blastall($sequences);
>   while(my $sbjct = $report->nextSbjct) {
>     while (my $hsp = $sbjct->nextHSP) {
>
>       # How can I find out which of the sequences in
>       # @$sequences caused the hit?
>
>       print "some sequence matched at ", $hsp->query->start,"->",
> 	 $hsp->query->end, "\n";
>     }
>   }
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l@bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-l
>

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