[Bioperl-l] EST functional categories

[Eckhard Lehmann]

> I'm working on an EST project.  I'd like to use bioperl to identify 
> putative functional categories of individual ESTs and eventually end up 
> with those ubiquitous pie charts or histograms (using excel).  I've seen 
> that ontologies structures have been built into bioperl but there isn't 
> much documentation (or I've overlooked it).  I could use some help/example

I am not quite sure about that - but isn't it better to do functional
analysis on contigs that are assembled of EST's, rather that individual EST's?

> scripts to get started.  I've blasted all my ESTs against nr.  How do I 
> take the best blast hit and look it it's functional category on
> GO/Interpro?

In the EST project that I am working on - I have the impression yet, that
you can not pull out much usefull information of these short sequence parts.
When it comes to BLASTx for instance, you get many many different confusing and
discusting non-informative hits.
With contigs you have mostly eliminated the redundancy and can do better
BLASTx. Functional analysis should then be possible by backtracking the EST
source information.

I have done some stuff, but it is mainly database oriented (PostgreSQL,
PL/PgSQL) and has not much to do with BioPerl currently. Used only BioPerl-DB to
load the EST sequences into a BioSQL setup along with contigs and things, and
now to fetch individual sequences into flat files sometimes.
However, I used Perl and GD::Graph for making pie charts and a web interface
to the database.

Eckhard ;)