Closed matentzn closed 5 years ago
Test the wbbt.obo to see where this gets you with your ingest errors! Let me know if there are any more.
There are unionof (union_of) relations got filtered out from the simple file. They should be retained. Just a note, unionof in WBbt is the same as owl2 DisjointUnion https://www.w3.org/TR/owl2-new-features/.
@raymond91125 I will look into this! Out of curiosity, what are you using union_of for?
some worm cells can come from one of two possible lineages (ancestry). thus we need some sort of XOR relationship.
can this wait till Friday? I think I need to have some discussions about doing this properly with my boss, who is on travelling at the moment..
Sure thing. Just note that I should really fix this type of relations. In Protege, presently, they are represented as (e.g.) 'P1.aaaa nucleus' or 'W.aaa nucleus' But they should be: ('P1.aaaa nucleus' or 'W.aaa nucleus') and not ('P1.aaaa nucleus' and 'W.aaa nucleus') I don't think either of these can be reasoned in Protege, though. HermiT seems to go on forever.
Contact me so I don't forget about this. I don't think your use case/biology maps onto OWL axioms in exactly the way it may seem on the surface.
We already took the discussion offline. we are getting rid of the logical axiom for now; see email thread.
@raymond91125 Sorry for the back and forth here. After some recommendations offline by @cmungall I decided to leave your modelling as it was for now and redefine the OBO export of wbbt as follows: Take the full version of wbbt, remove all non-native (non WBbt) classes, remove redundant subsumptions and export to OBO. Please test wbbt.obo one last time here. I hope the typedefs in the end are not bothering you, else I will get rid of them in a second step.
@raymond91125 Have you had a chance to look at this?