[Bioperl-l] Re: bioperl interface to EMBOSS

Jason Stajich jason@cgt.mc.duke.edu
Mon, 4 Mar 2002 17:20:46 -0500 (EST)


Michael -

Are you running CVS version of the code?  This functionality is only in
the live version of the code until 1.0alpha2 goes out this week.

The place to put new code is in Bio::Tools::Run::EMBOSSApplication in the
run method.  I went ahead and added code that writes out SimpleAlign
objects in addition to Seq objects.  You can pick that up from CVS.  Let
me know if you are still having trouble.  I added a test for this
functionality in t/EMBOSS.t so you can see how it is called.

Beginning to think this bit of logic should get factored out into a
factory at some point so that it can be used by other programs which want
bioperl objects to be dumped into files for use in apps.

FYI - you can generate a rough consensus sequence with the SimpleAlign
interface $aln->consensus_string(50) where 50 is bases the occur 50% or
more being chosen.  This may remove the need for cons altogether depending
on whether or not you are using DNA or Protein data.  We could probably
extend our consensus string method to include ambiguity symbols and what
not, but for now it is just implementing a simple "most votes wins, not
enough votes gets an X".

-jason
On Mon, 4 Mar 2002, Muratet, Michael wrote:

> Jason
>
> I'm trying to run EMBOSS/cons without creating the file. Back in February
> you talked about some code you wrote for the water routine. I don't see in
> the documentation how to extend that. Can it be extended? Thanks.
>
> Mike
>

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