[Bioperl-l] Bio::Tools::Run::StandAloneBlast.pm - bl2seq question

[Chris Fields]

Wouldn't surprise me too much if the parameters are not set the same; IIRC the main BLAST URL API and the online NCBI Web-BLAST have different default settings.

chris

On Aug 18, 2010, at 5:32 PM, Dave Messina wrote:

> Hi Caleb,
> 
> The first thing I would do is take BioPerl out of the equation and test your local bl2seq on the command line. If you get the same output locally as on the web version, then there is a problem with BioPerl. If you're still seeing a discrepancy between the web and your local run, then this isn't a problem with BioPerl.
> 
> Just to be clear, BioPerl doesn't "implement" any of the BLAST programs; it is simply a wrapper around the programs that you download from NCBI. That doesn't mean BioPerl isn't at fault, of course, just that it's important to isolate the problem carefully.
> 
> The most common reasons for these discrepancies are:
> 
> - different version numbers of BLAST
> 
> 2.2.21? 2.2.22? Is it the same on the web as locally?
> 
> - similarly, different implementations of BLAST
> 
> NCBI's old BLAST suite is now deprecated and replaced with BLAST+. All of the online BLAST web queries are Blast+ now — are you running BLAST+ locally? (there's also a separate BioPerl wrapper for BLAST+ called Bio::Tools::Run::BlastPlus)
> 
> - hidden "default" parameters
> 
> Even though you're only changing a handful of parameters, the defaults (particularly on the web version) may be different than what you expect.
> 
> In your case, it looks like on the web version, match score is 2 and mismatch is -3. However, in the local version I believe match score is 1 and a mismatch is -3.
> 
> See this line in the params block near the end of your post:
> 
> 	'matrix' => 'blastn matrix:1 -3',
> 
> 
> 
> Dave
> 
> 
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioperl-l