Replace CARO IDs in cob-non-native OWL export

[cmungall]

Background

There are two COB products, one with native COB IDs, the other with OBO IDs rewired in.

Please do not use this issue to discuss whether other OBO IDs should be rewired in, or which of the two products should be primary. You are welcome to discuss this here:

• Release COB with COB ids + mappings, rather than with the mapped IDs #244

The subject of this issue is what to do with the rewiring to CARO

Currently there are 3 IDs rewired to CARO:

Their mappings are here:

COB/src/ontology/components/cob-to-external.tsv

Line 58 in 9cdc2c4

COB:0000021 gross anatomical part owl:equivalentClass CARO:0001008 gross anatomical part . semapv:ManualMappingCuration

COB/src/ontology/components/cob-to-external.tsv

Line 63 in 9cdc2c4

COB:0000022 organism owl:equivalentClass CARO:0001010 organism or virus or viroid . semapv:ManualMappingCuration

COB/src/ontology/components/cob-to-external.tsv

Line 112 in 9cdc2c4

COB:0000118 cellular organism owl:equivalentClass CARO:0010004 cellular organism . semapv:ManualMappingCuration

Recently developers of various anatomy ontologies decided to obsolete CARO in favor of Uberon. The Uberon class for e.g "tissue" would no longer mean "animal tissue", but simply "tissue". Ontologies like GO have swapped out CARO for Uberon, but this process is not complete for OBO as a whole.

For the full discussion, see:

• Clarify CARO-Uberon connections obophenotype/uberon#2349

This change has not been reflected in cob-to-external. For example, Uberon immaterial anatomical entity is a subclass of the COB class of the same name:

COB/src/ontology/components/cob-to-external.tsv

Line 77 in 9cdc2c4

COB:0000056 immaterial anatomical entity sssom:superClassOf UBERON:0000466 immaterial anatomical entity . semapv:ManualMappingCuration

Leaving CARO in the rewired COB is problematic, since it is being phased out.

Replacements

These could potentially be broken into subtickets. I propose the following:

1. CARO:0001008 ! gross anatomical part ==> add to uberon (NTRs: gross anatomical part, animal gross anatomical part obophenotype/uberon#2921) and use this
2. CARO:0001010 ! organism ==> do NOT use NCBITaxon:1, see Propose changing the label on the NCBI top level term to 'NCBI root' obophenotype/ncbitaxon#10. Use COB:0000022
3. CARO:0010004 ! cellular organism ==> NCBITaxon:131567 "cellular organisms"
4. COB:0000056 ! immaterial anatomical entity ==> use Uberon (by switching predicate in mapping)

Note I am including 4 in this as well, it looks like the CARO mapping was accidentally omitted

I think 3 is a little unsatisfying due to the pluralization. However, this is consistent with the policy of using a native OBO ID in the rewired product

[ddooley]

So COB:0000022 organism stands as the new ID for organism going forward? This is because it encompasses more than OBI Organism (i.e. viruses), and less than NCBITaxon (=not quite an ontology) root? UBERON doesn't want to mint an "organism" term because viruses don't fall within purview of UBERON's anatomy domain?!

[bpeters42]

I am not sure what this discussion is about or where it came from OBI:0100026 (=organism) very much includes viruses https://www.ebi.ac.uk/ols4/ontologies/obi/classes/http%253A%252F%252Fpurl.obolibrary.org%252Fobo%252FOBI_0100026 It is the most widely re-used class from OBI in the OBO world. But it is not in OBI's scope. So we wanted to cede it to COB, but it should have the exact meaning that OBI:0100026 has, so that it can be painlessly replaced by anyone using the current class. - Bjoern

…

On Thu, Jan 25, 2024 at 1:30 PM Damion Dooley ***@***.***> wrote: So *COB:0000022 organism* stands as the new ID for organism going forward? This is because it encompasses more than OBI Organism (i.e. viruses), and less than NCBITaxon (=not quite an ontology) root. UBERON doesn't want to mint an "organism" term because viruses don't fall within purview of UBERON's anatomy domain?! — Reply to this email directly, view it on GitHub <#248 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADJX2IQ2TX3LPEU57BUNA6TYQLFJHAVCNFSM6AAAAABCJV2G5WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMJRGAZDSNJTGI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Bjoern Peters Professor La Jolla Institute for Immunology 9420 Athena Circle La Jolla, CA 92037, USA Tel: 858/752-6914 Fax: 858/752-6987 http://www.liai.org/pages/faculty-peters

[cmungall]

Please let's keep the discussion focused on the issue at hand, @bpeters42 is right, this has nothing to do with viruses, these are included, this has already been decided, if you want to relitigate keep it on this issue, thanks!!

• Integrate 'organism' #6

[ddooley]

Ok, I misstepped there - not sure where I got notion that OBI organism didn't include viruses. But my point was that above @cmungall you said "Use COB:0000022". I was trying to figure out why you didn't say "use OBI:0100026"? If COB:0000022 maps to OBI:0100026 then I understand.

[cmungall]

np. Not sure I understand though - does OBI really want to be the owner of a general concept like "organism"?

…

On Wed, Jan 31, 2024 at 4:28 AM Damion Dooley ***@***.***> wrote: Ok, I misstepped there - not sure where I got notion that OBI organism didn't include viruses. But my point was that above @cmungall <https://github.com/cmungall> you said "Use COB:0000022". I was trying to figure out why you didn't say "use OBI:0100026"? If COB:0000022 maps to OBI:0100026 then I understand. — Reply to this email directly, view it on GitHub <#248 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAMMOMGJMUJCUS2KXX3243YRI2GZAVCNFSM6AAAAABCJV2G5WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMJZGAYTCNJWGA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jamesaoverton]

OBI does not want to own 'organism'. Maybe @ddooley is thinking about the perennial question of COB "adopting" the OBI:0100026 identifier vs using the new COB:0000022.

[ddooley]

Yes! Am I to use COB:0000022 in FoodOn or ONS or whatever. Or use OBI:0100026? This ID that I'm specifying in my COB import file.

[cmungall]

COB:0000022 is what you should use moving forward. Of course, it won't resolve until this issue is closed.