surface structure is an asserted parent of cornea - coming from imports

[nicolevasilevsky]

surface structure is an asserted parent of cornea but this has been removed as an asserted parent in monarch-initiative/mondo#542. This assertion is coming from the NBO and PATO import.

This is causing an issue in Mondo: monarch-initiative/mondo#542

Thanks!

[matentzn]

Not sure I understand - how is this an uberon issue if it comes from the PATO import? Is PATO simply out of sync with uberon? If this issue is still not fixed Uberon, you could simply fix it yourself!

[nicolevasilevsky]

Not sure - there are two asserted superclasses, one from NBO and one from PATO.
Can I refresh the PATO and NBO imports?

Otherwise, how do I fix it myself?

Thanks! :)

[matentzn]

Try doing a PATO release, and review wether the offending change gets removed. If it does, you can do the same for NBO - this needs to get a release anyways.. I am there to advice but I will try not do the actual work anymore - not because I am lazy (you know I love this stuff) but because I need to make myself redundant!

[nicolevasilevsky]

sure, I'll try to run a PATO release now :)

[nicolevasilevsky]

actually, looking into it further - PATO has been released since this issue has come up and it looks like it is not fixed:
https://github.com/pato-ontology/pato

[matentzn]

Did you check if PATO.owl has the offending axiom? If not, maybe you have to refresh the PATO import in Mondo?

[nicolevasilevsky]

I think this is an issue in PATO

Should I remove this assertion in PATO, then do a release?

[balhoff]

@nicolevasilevsky you are looking at a very old PATO. I don't see that in the current release. I do see this in NBO, which probably needs its Uberon import refreshed.

[nicolevasilevsky]

Ah, okay, I'll try to do that. Thanks @balhoff!

[nicolevasilevsky]

I updated the Uberon import but I still see surface structure asserted in NBO

[balhoff]

Sorry, I just actually looked, it is still in Uberon! I thought it was removed in one of those other PRs: #2070 ??

[balhoff]

It's coming back into Uberon via its own pato and nbo imports. I guess that's exactly what you started off saying here. Sorry! We need base files. 🥴

[nicolevasilevsky]

Is there anything I can do to help? I assume the need for base files is on the radar?

[nicolevasilevsky]

I can close this if there is nothing actionable at the moment

[balhoff]

It's a self-perpetuating axiom that lives entirely within a recursive import circle. I think a good place to start would be to change the Uberon build to at least filter external axioms from the PATO and NBO downloads, using something like this: https://github.com/INCATools/ubergraph/blob/a27eda415f2379ffdcdc563feaa466c0e3a16830/Makefile#L19-L25

[nicolevasilevsky]

@balhoff or @matentzn could either of you help me with this? thank you!

[nicolevasilevsky]

we discussed this on the Uberon call on 01/24/22.

This may be a makefile issue for Uberon.

once the NBO base file is added per obo-behavior/behavior-ontology#117, we then need to change the makefile configuration in Uberon such as the NBO and PATO base files are used to make the imports in Uberon.

@shawntanzk (thank you!)

[matentzn]

This issue will be solved by moving Uberon to base. Priority.

[shawntanzk]

@nicolevasilevsky - I'm going to close this ticket (given we moved nbo and pato to base file). We will make a UBERON release this week. If there is still any issue, please reopen. Thanks!