[Bioperl-l] project ideas

[Hilmar Lapp]

Brian,

this really is a non-trivial task, because essentially what we did is 
molding an LL entry into a Bio::Seq, which to be quite honest leaves 
the result with more than just a cone head. The fundamental problem is 
that Bio::SeqI really is about one sequence, whereas an LL entry is 
more like a gene model, with different products each representing a 
splice variant etc.

It doesn't fit a SeqFeature::Gene::GeneI either though, since it's not 
a feature with sub-features. Nor do you really want to mold it into a 
ClusterI I think.

So, to really solve this properly IMHO you'd have to create a class 
that better reflects the data model behind these gene-centric entries. 
Maybe something along the lines of a Bio::GeneI or Bio::GeneModelI, 
which AFAIR surfaced repeatedly over the past years ... (but no-one 
ever got serious about it).

Could actually be something to chat about in Singapore.

	-hilmar

On Wednesday, February 12, 2003, at 06:01  AM, Brian Osborne wrote:

> Richard,
>
> There's one SeqIO module that's had a good start but is unfinished,
> SeqIO::locuslink. What it's missing is a hierarchical treatment of the
> LocusLink fields - currently it returns all the values, I think, but in
> flattened form. For example, a gene may have multiple "products" or
> "transcript variants" but the association between these sequences and 
> their
> respective attributes, like domains for example, is lost. Its original
> author knows this but has said he doesn't currently have time to 
> complete
> the module.
>
> No, this is not a test, but I'd say it's an important module even if 
> it's
> only a "Genbank thing".
>
> Brian O.
>
> -----Original Message-----
> From: bioperl-l-bounces at bioperl.org 
> [mailto:bioperl-l-bounces at bioperl.org]On
> Behalf Of Richard Adams
> Sent: Wednesday, February 12, 2003 8:41 AM
> To: bioperl-l at bioperl.org
> Subject: [Bioperl-l] project ideas
>
> I'd like to start contributing to bioperl so if it's a good place to
> start I'll volunteer
> for designing new Bio::SeqIO test cases, if it's still going.
> Let me know what sort of tests you'd need.
>
> Richard Adams
> Universitry of Edinburgh
> UK
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-l
>
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-l
>
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------