search

evoinfo / cdao

Comparative Data Analysis Ontology - A formalization of concepts and relations relevant to evolutionary comparative analysis
Creative Commons Zero v1.0 Universal
9 stars 4 forks source link

Updates to improve status in OBO dashboard #20

Closed balhoff closed 7 months ago

balhoff commented 7 months ago
hlapp commented 7 months ago

Thanks for this @balhoff. Can we in the future agree that all PRs require review before being merged? There are a lot of dependencies on CDAO (for good or bad).

balhoff commented 7 months ago

@hlapp, absolutely, sorry about that. I wasn't sure there were others paying attention here. Please let me know if you see anything to reconsider in this PR.

hlapp commented 7 months ago

I think the two non-trivial changes here are (a) changing the identifier (IRI) of the property under which to find a definition, and (b) re-axiomatizing the has property.

Change (a) could throw off some custom-built UI, but I don't know about any that are used in the context of CDAO. So I think this should nonetheless be inconsequential at the practical level.

Change (b) changes (at least potentially) reasoning results, so I have more reservations about it. Is this change necessary for some use-case, or motivated by other reasons? Can there be a public feedback period? So in short, if it's reasonably possible to turn this into its own issue where a rationale is given and potential consequences outlined, with an associated PR, that'd be ideal. But if there's already dependency on that change somewhere, I guess we can also leave the change in place, Though even in that case, it'd still be good to have this change documented and referenceable in an issue for later.

balhoff commented 7 months ago

@hlapp there should not be any reasoning changes, since CDAO_0000194 and CDAO_0000190 have been previously asserted as equivalent properties (and still are). Since I marked CDAO_0000194 as deprecated, I wanted to make sure all usages in axioms were in terms of CDAO_0000190 rather than CDAO_0000194. I can turn this into an issue.