Re: [CDAO-discuss] [Phenoscape] New page for discussion about top down inferences

[Hilmar L. <hl...@ne...>]

On Dec 8, 2009, at 2:26 PM, Jim Balhoff wrote:

> On Dec 8, 2009, at 2:14 PM, Hilmar Lapp wrote:
>
>> On Dec 8, 2009, at 2:00 PM, Jim Balhoff wrote:
>>
>>> The relationship is asserted via the matrix, so we have not  
>>> specified the link anywhere so far.  I suppose we could add  
>>> metadata to each cell stating what relationship to use between the  
>>> taxon and the state.
>>
>> So there must be a relationship according to spec, and in order to  
>> meet RDFa compliance. To what is the phenotype related right now  
>> (the state? character?), and using which relationship? Or is that  
>> not how it works right now?
>
> The phenotype is related to the state via  
> "phenoscape:describesPhenotype".  In this way it is RDFa compliant,  
> but there is no RDF triple which can be extracted relating the  
> phenotype to any taxon.  This must be created by a NeXML-aware  
> application, as this is what the matrix is there for.

Could a CDAO-aware application reason though that if taxonA hasState  
state1 and state1 describes:Phenotype phenotypeA then taxonA exhibits  
phenotypeA?

If not, we should work on getting there. I'm copying the CDAO list  
here as I know that they are very excited about the Phenoscape use- 
cases for applying CDAO.

In the meantime, can we stick the phenotype in NeXML as its own  
"think" that gets an ID, and then relate it to the state *and* the  
taxon using the ID and the applicable relationships? (I'm also copying  
the NeXML list here in case they can help with answering that.)

	-hilmar
-- 
===========================================================
: Hilmar Lapp  -:- Durham, NC -:- informatics.nescent.org :
===========================================================