[BioSQL-l] SeqFeature scores

[Chris Fields]

On Oct 16, 2008, at 5:12 PM, Hilmar Lapp wrote:

>
> On Oct 14, 2008, at 1:28 PM, Chris Fields wrote:
>
>> Personally I'm not sure how we'd store score data in BioSQL.  Is  
>> 'score' within the schema?  I suppose we could add it as a specific  
>> tag value but that seems potentially hackish and prone to naming  
>> conflicts.
>
> Well, yes there might be naming conflicts, if someone wanted to add  
> a tag to a seqfeature called 'score', for example.
>
> But would there be cases where it would make sense to have value  
> stored in the feature's $feat->score() method *and* a (semantically)  
> different one as the 'score' tag's value? Quite frankly I'd be hard- 
> pressed to come up with a scenario where that might make sense.

Agreed.  I think it's a very unlikely scenario, frankly, but never  
hurts to bring it up.

> So as far as I am concerned I wouldn't actually have a problem with  
> changing the implementation of score() to store/pull the value to/ 
> from the tag/value hash. In fact, that's what B::SF::Similarity does  
> for the attributes it adds methods for (such as bits, significance,  
> etc).
>
> Thoughts? I'm copying this to the Bioperl list as really it is a  
> BioPerl/Bioperl-db issue.
>
> 	-hilmar

Makes sense to me.  I think anything else not supported in BioSQL/ 
bioperl-db should likewise be stored in the tag/value hash, but  
score() is the only one that comes to mind at the moment.  I'll
make the change once the bug report is filed so we can track any  
problems I encounter.

chris