[Ontology-editors] Cellular component 2.0

[Chris Mungall]

When I was in Cambridge we started discussing the idea of adding new  
relations to the cell component ontology.

I found I also needed these kinds of relations when generating  
necessary and sufficient conditions for CC terms in:

http://wiki.geneontology.org/index.php/XP:cellular_component_xp_self

These can be browsed in oboedit, but perhaps the best way is via the  
html table:
http://wiki.geneontology.org/index.php/XP:cellular_component_xp_self#availability

I think having computable definitions is extremely valuable for GO. We  
can do so much more graph management automatically. For downstream  
consumers the additional relations should prove useful too.

Where should we take this next? This is in fact much simpler than the  
external cross-products. There is nothing to stop us selectively  
vetting and moving these xp defs into the main gene_ontology_write.obo  
as soon as OE2 is released. Of course we would want to give the  
appropriate notice to users. We can of course keep editing  
cellular_component_xp_self.obo as a separate file too, but it would be  
good to have an integration plan

Cheers
Chris