[Bioperl-l] Bio::DB::GFF::Aggregator problem, new wormbase models.

[Lincoln Stein]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'll make a new prepackaged aggregator for this as soon as WormBase 
makes the final and long-anticipated transition to real genes.

Lincoln

On Wednesday 18 February 2004 12:10 am, Todd Harris wrote:
> Hi Phillip -
>
> You need to aggregate the separate parts of the CDS.  Create a
> wormbase_cds (or whatever you wish to call it), aggregating the
> following features using the CDS group: coding_exon,5_UTR,3_UTR.
>
> The following stanza should do the trick.
>
> $dbgff = (-adaptor => 'dbi::mysql',
>           -dsn     =>
> 'dbi:mysql:database=your_database;host=your_host', -aggregators =>
> [qw(wormabse_cds{coding_exon,5_UTR,3_UTR/CDS})], -user    =>
> 'your_username',
>           -pass    => 'your_dbgff_pass');
>
> This should do the trick for properly aggregating genes under the
> new WormBase CDS class.
>
> Todd Harris
>
> > On 2/17/04 2:18 PM, Philip MacMenamin wrote:
> >
> > I cannot get things to aggregate properly in the new wormbase
> > models. Previously this worked:
> > $aggregator = Bio::DB::GFF::Aggregator->new(-method    => 'test',
> >       -sub_parts => ['UTR','CDS:curated']
> >      );
> > (Or I could of more simply used the 'processed_transcript'
> > aggregator that Linoln wrote.)
> >
> > And one feature, with the UTRs "aggregated" together with the
> > curated CDS's, was returned:
> > test:5_UTR(AH6.5), start->9524078 end->9526248
> > This could be fed to some kind of glyph, and would draw a nice
> > picture of UTRs hanging off a coding seq.
> >
> > [--------------]    [----------]  [---->
> > (You have to use your imaginatio a bit here)
> >
> > With the new models of wormbase, this is not the case, so I am
> > re-writing code to accomadate these changes.
> >
> > Now I am returned for example 3 features: a 3 prime UTR, a prime,
> > and a CDS. test:UTR(5_UTR:AH6.5)9524078 9524086
> > test:UTR(3_UTR:AH6.5)9525782 9526248
> > test:curated(AH6.5)9524087 9525781
> >
> > The glyph then draws these things on two planes.
> > []                [------->
> > [---------]  [------------] [----------]
> >
> > My assumption is that a single feature can be rendered as a
> > single glyph. And if you have >1 feature, then >1 glyphs are
> > needed. Therefore, I assume the problem stems from my failure to
> > aggregate UTR and CDS features.
> >
> > I have noticed that the new SQL GFF Db used to have two fmethods
> > for UTRs, one for both 3 and 5 primes. The new one has only one.
> > There are other changes of course, but I think this is important.
> >
> > So, essentially I want to draw UTRs with CDS's attached... with
> > the new wormbase models. If anyone knows how to do this, i'd like
> > to hear from them.
> >
> > Thanks :)
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at portal.open-bio.org
> http://portal.open-bio.org/mailman/listinfo/bioperl-l

- -- 
Lincoln D. Stein
Cold Spring Harbor Laboratory
1 Bungtown Road
Cold Spring Harbor, NY 11724
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQFAMy8h0CIvUP7P+AkRAtCSAJwPVPdXqs9rXSFYdCD8lVhsB/5wkACdEOrr
EwcJXiat61tP3F5XJXA1j+c=
=P0TB
-----END PGP SIGNATURE-----