[BioPython] pblast in NCBI's website differs from biopython's pblast

[Peter]

On Wed, May 28, 2008 at 4:41 PM, Raul Guerra <colochera at gmail.com> wrote:
> Hi everyone,
>
> I was wondering if someone has had the same problem. I am running the
> following code in BioPython. ...
>
> The results that I got back are different from the results I get from the
> pblast option  at http://www.ncbi.nlm.nih.gov/blast/Blast.cgi (I refer to it
> when I mention pblast in NCBI's website) ...
>
> I also know that the pBlast in NCBI's website utilizes a Gap Cost of
> "Existence: 11 Extension:1". I am not sure how to translate that into the
> qblast function in Biopython. I am not sure if this is the problem, but it
> could be that Biopython's pblast and NCBI's pblast have different
> parameters.

The NCBI have often changed their default options in the BLAST tools,
both the online versions and the web interface.  As you have seen from
the Biopython source code, we set many of the options to specific
default values which may no longer match what the NCBI webpages do by
default.  Its a little frustrating :(

For a similar examples from last year, see
http://portal.open-bio.org/pipermail/biopython/2007-August/003679.html
and http://portal.open-bio.org/pipermail/biopython/2007-August/003693.html

You should compare the parameters for the two sets of results, work
out where they are different, and decide which settings are best for
your problem.

For the gap costs, see
http://www.ncbi.nlm.nih.gov/BLAST/Doc/node28.html - you can specify
this parameter in the Biopython qblast function with the optional
gapcosts argument.

Peter