[Biopython] Pulling Alignment From PSI-Blast Output

[Peter]

On Tue, Feb 8, 2011 at 5:48 PM, Brett Bowman <bnbowman at gmail.com> wrote:
> Sadly no - I tried lining up the output sequence alignments, but the result
> is meaningless because they are all just aligned pair-wise to the query.
>  I'm wondering if maybe I just need to go back and use BlastPGP somehow?  I
> know they cut out a lot of features to make the PSIblast standalone
> executable.  Though why they would remove things from the output makes no
> sense to me...
>
> So I guess that goes back to my previous question - if parsing the PSIBlast
> XML output only gives me a Bio.Blast.Record.Blast object, then where do the
> Bio.Blast.Record.PSIBlast objects, which are supposed to have that alignment
> built in, come from?

Currently you've have to use the old text based PSI-BLAST parser to
get Bio.Blast.Record.PSIBlast objects. The XML parser doesn't (yet) do
anything special for PSI-BLAST. But in any case the PSI-BLAST XML
just has the pairwise alignments (right?).

Peter