[Bioperl-l] GO annotatinos in BioPerl. A tentative proposal...

Ewan Birney birney@ebi.ac.uk
Thu, 5 Apr 2001 19:50:16 +0100 (BST)

On Thu, 5 Apr 2001, Hilmar Lapp wrote:

> Ewan Birney wrote:
> > 
> > <ewan's prejudice>
> > Annotation objects always seem to degnerate into something like name-value
> > pairs/multiple mix-in inheritance of interfaces. It is way too slippery to
> > pin down...
> > </prejudice>
> Hm. You mean there is no sensible AnnotationI a generic
> AnnotationDumper (like one dumping to GFF format) could be
> satisfied with? I of course don't know the past experiences with
> any attempts in this direction. (Does anyone have the BioJava
> solution at hand? Is there a IDL definition [BioCorba, BSA, ...]
> making an attempt at this?)

I think the BioJava annotation is just name-value pairs (with some extra
funky ability to store "nulls" in values). 

Bioperl annotation currently is

   A collection of DBLinks, A collection of References and A collection of

This is not such a bad grouping on reflection...

Notice that GO could just be added as another DBLink type object. Not
quite what mark wants.

Anyway - back to your question - there is no obivous "flat file format" or
model that clearly encompasses a common core of "annotation". There is
always a need for sort of semi-structured extensions.

Not sure where this leaves us.

> 	Hilmar
> -- 
> -----------------------------------------------------------------
> Hilmar Lapp                              email: hilmarl@yahoo.com
> GNF, San Diego, Ca. 92122                phone: +1 858 812 1757
> -----------------------------------------------------------------

Ewan Birney. Mobile: +44 (0)7970 151230, Work: +44 1223 494420