search

geneontology / go-ontology

Source ontology files for the Gene Ontology
http://geneontology.org/page/download-ontology
Creative Commons Attribution 4.0 International
216 stars 39 forks source link

Review subclass relations remaining/missing after regulates-part-of changes #20069

Closed balhoff closed 3 years ago

balhoff commented 3 years ago

We removed the regulates o part_of property chains in #18653.

Relationships (possibly indirect) which you were not removed by deleting the property chain, but which you may want to remove from the ontology somehow:

Relationships (possibly indirect) which you were removed by deleting the property chain, but which you may want to add back somehow (either directly or by adding a regulates relation):

pgaudet commented 3 years ago

Hi @ValWood

We used to have this relation: GO_0090235 regulation of metaphase plate congression is_a GO_0010564 regulation of cell cycle process but in the non-regulatory branch of the ontology there is no link between metaphase plate congression and cell cycle process ; should there be ? (I see it in QuickGO but not in the ontology).

Thanks, Pascale

ValWood commented 3 years ago

Yes, metaphase plate congression is a cell cycle process.

pgaudet commented 3 years ago

ok thanks

pgaudet commented 3 years ago

Hi @ValWood

regulation of mitotic nuclear envelope disassembly is_a regulation of mitotic nuclear division This is because 'mitotic nuclear envelope disassembly' is a 'mitotic nuclear division' - should it not be part of ?

Thanks, Pascale

ValWood commented 3 years ago

I think part_of yes

pgaudet commented 3 years ago

@ValWood Are you OK with me removing GO:0051784 negative regulation of nuclear division is_a 'negative regulation of organelle organization' ?

ValWood commented 3 years ago

I guess so.

pgaudet commented 3 years ago

I guess so.

I dont want to bully you into saying you ! If you think this relation was appropriate I'll keep it.

ValWood commented 3 years ago

No it isn't that. Is it causing problems?

I feel a bit bad because I don't really know if the particular annotation should be part_of, regulation, or causally upstream. I have a note to try to pin down the starts and ends of the process. So I was being deliberately vague! I was hoping I would have a better clue from some upcoming papers and asking the authors but I didn't get round to it yet... So you can remove it...regulation of organelle organization isn't very useful...

pgaudet commented 3 years ago

I understand what you mean.

But more generally, organelle division and organelle organization should be disjoint processes (even if some components may be overlapping?)

pgaudet commented 3 years ago

I'll also remove

@ValWood unless you think those should remain

pgaudet commented 3 years ago

@ValWood

Is it causing problems?

Not really, just those are 'leftover' from the removal of the regulates-part-of inference chain.

According to @balhoff under regulation, all subclasses should be inferred based on the relation in the non-regulation part of the ontology. So for example if we decide 'nuclear envelope disassembly' is_a 'organelle organization' then this will be inferred in the regulation branch.

However if it's a part_of, it should not become a is_a in the regulation branch. (Jim, correct me if this is not right).

Does that seem reasonable ?

ValWood commented 3 years ago

OK agreed!

pgaudet commented 3 years ago

I will not fix the others - for example one suggestion was to add

regulation of neurotransmitter secretion is_a regulation of cellular localization

but in the non-regulation branch, serotonin uptake is_a establishment of localization which is part_of cellular localization

Therefore regulation of uptake is not a regulation of localization. This is consistent.