[Bioperl-l] Bio::SeqFeature::Annotated API change

[Chris Fields]

On Oct 12, 2007, at 9:46 AM, Hilmar Lapp wrote:

>
> On Oct 12, 2007, at 9:04 AM, Sendu Bala wrote:
>
>> Or, heck, since it's such a mess, do we just want to drop Annotated
>> entirely and recommend using AnnotationAdaptor with Generic? What  
>> does
>> Annotated bring to the table?
>
> My recollection may be wrong since it's been quite a while, but I  
> believe one of the main motivations was to reflect the ontology- 
> typing of feature type and other annotation that GFF3 requires.
>
> The motivation of AnnotationAdaptor was primarily to provide a view  
> onto a SeqFeatureI that makes tag/value annotation transparently  
> appear as B:AnnotatableI and B:AnnotationCollectionI compliant,  
> mixed in with the annotation collection that B:S:Generic may hold  
> already.
>
> Although sometimes confused, these are not the same motivations.  
> AnnotationAdaptor provides a view - you use it when that's the view  
> you need on a SeqFeatureI object you have in hand. For example,  
> bioperl-db uses it so it doesn't have to bother about how to de/ 
> serialize tag/value pairs, when it knows how to de/serialize  
> annotation collections already from sequence objects.  
> AnnotationAdaptor doesn't care whether the primary_tag, or any  
> other tag, is from an ontology.

Ah, forgot it returns a Collection.  It doesn't modify the SF I assume?

>> (Since its one of the modules added in 1.5 I think it's fair to  
>> drop it
>>> from the 1.6 release: stuff that doesn't work can and should be  
>>> moved
>> out of the path of stable branches.)
>
> I agree with that unless someone steps up and makes it work.
>
> 	-hilmar

Making it work may be more trouble than it's worth.  Sincerest  
apologies to the authors of the module but we should probably  
deprecate it in favor of something more flexible when we can.  It's  
obvious anything reliant on it would also have to be dropped.

The GMOD meeting is a few weeks away, where I assume some things will  
be hammered out re: GFF3 and BioPerl.  The idea of FeatureIO is worth  
saving; maybe retooling it would be in our best interests, but it  
needs a decent roadmap to move forward.

chris