[Bioclusters] mpiBLAST Performance

[Jeremy Mann]

>
> I forgot to mention, I'd estimate mpiBLAST is at least 50% faster than
> my homebrewn
> script which just subdivides the blast query and searches a copy of the
> database on
> each node.  Being able to put database fragments entirely (or mostly) in
> RAM obviously
> helps.  I'm still wondering though why mpiblast doesn't assign each node
> a specific
> piece of the database in local storage.  I am looking at the local
> storage area on one
> of my nodes (n2) and there is nr.00, nr.05 and nr.13 indices.  Should
> each node just get
> one piece?  Or does every node evetually get the entire thing as you run
> it over and
> over again?

I actually asked these same questions about two weeks ago. You would
*think* each node would have its own segment of the database.


-- 
Jeremy Mann
jeremy@biochem.uthscsa.edu

University of Texas Health Science Center
Bioinformatics Core Facility
http://www.bioinformatics.uthscsa.edu
Phone: (210) 567-2672