OBO challenge 2023: go green in CL OBO dashboard

[paolaroncaglia]

Placeholder ticket inspired by @matentzn 's talk at ICBO 2021. CL does quite well in the OBO dashboard, but we could address those missing definitions between now and 2023 :-)

(BTW @matentzn @dosumis the CL report says 397 missing defs, but the yellow table only shows 200, and I can't see a link to go to a second page? If I download the TSV, it's also limited to 200 entries. I see the same issue in a couple of other ontologies' reports that I checked. Let me know if I should file a bug ticket in the OBO tracker instead.)

[matentzn]

I love it thanks so much @paolaroncaglia

[paolaroncaglia]

I'll self-assign for shepherding purposes. Quick thoughts:

• [ ] CL terms that lack definitions may be grouped by broader classes (e.g. neuron types), to facilitate working on related terms in bulk if desided, possibly by editors working on related project;
• [ ] Some of this work may be a good first task for novel CL editors, @dosumis if you think it'd be appropriate for the future Cell Annotation Platform (CAP) developer.

[paolaroncaglia]

Note to self: the OBO report (cl-edit.owl-obo-report.tsv) lives here and is updated at each release. Still 372 warnings for CL missing definitions.

[paolaroncaglia]

Following up on the 372 CL terms that lack a textual definition: I looked at the first 5 CL terms (sorted by ascending ID as per cl-edit.owl-obo-report.tsv). 3/5 have an equivalence axiom, e.g.

'neuronal receptor cell' Equivalent To 'sensory receptor cell' and neuron and ('capable of' some 'detection of stimulus involved in sensory perception')

Would it be possible to do an automated pass at generating textual defs based on equivalence logical defs, please? The results should be quite reliable (if perhaps in need of a quick manual revision to check for weird wording). I'll add this to the tech board.

[shawntanzk]

for definitions: if we implement patterns for #450 - a good portion of defs can be autogenerated using DOSDP for RO terms: @shawntanzk to check terms in http://dashboard.obofoundry.org/dashboard/cl/fp7.html and see how they are used. They should either be in RO or removed. @shawntanzk to make a proposal and bring it up again in a call or ticket

[github-actions[bot]]

This issue has not seen any activity in the past 6 months; it will be closed automatically in one year from now if no action is taken.

[ghost]

Checking in on this stale ticket... @shawntanzk and @paolaroncaglia / @paola-scibite, can this ticket be closed? If not, can the open action items be clearly listed?

[shawntanzk]

I'm happy for it to be closed - did what I can already. @matentzn pinging you just in case you want this kept open

[paola-scibite]

@bvarner-ebi @shawntanzk @@matentzn Thanks for following up. Atm the dashboard report says 369 missing CL definitions. I can't contribute, and I don't know what the status of available resources is at CL, so I'll be happy with whatever decision you take. Thanks!

[shawntanzk]

Just to add some context: The dashboard has no reds and the only things up are 4 non-RO properties and a whole bunch of missing text defs http://dashboard.obofoundry.org/dashboard/index.html the non-RO properties will likely not be solved - I think there is arguments for keeping some non-RO properties, and some are expansions etc. which are highly specific and will unlikely be accepted by RO. Missing defs wise, I'm guessing this are auto migrated stuff, and unless we have a way of programmatically pulling in the defs (maybe from xrefs or smth), I think its quite hard for someone to curated 369 defs. It is why I personally think we have gotten as far as we can with this ticket and it can be closed :) Hope that helps.

[ghost]

Thanks very much for the context, @shawntanzk. @matentzn, given the comments above, is it reasonable to close this issue? If there is no further feedback by next week, I propose to close the ticket and reopen if needed.

[matentzn]

I agree, awesome work everyone, this can be closed! 🎉