[ITR] Add GO to HCAO

[ami-day]

Hi @paolaroncaglia @zoependlington

We are working on updating the HCA metadata schema to include new ontology modules for:

• Treatment protocol method
• Treatment protocol target pathway

An example Treatment protocol method is: T cell activation GO:0042110 https://www.ebi.ac.uk/ols/ontologies/go/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FGO_0042110

An example Treatment protocol target pathway is: Positive regulation of memory T cell differentiation GO:0043382 https://www.ebi.ac.uk/ols/ontologies/go/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FGO_0043382

Would it be possible to add GO (Gene Ontology) to HCAO and also to ensure everything under OBI:0000070 (https://www.ebi.ac.uk/ols/ontologies/obi/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FOBI_0000070) is in HCAO as well?

Also open to feedback if you have any questions/suggestions!

Thanks!

[zoependlington]

Hi @ami-day, just to update you that Paola is no longer working on HCA. @bvarner-ebi will be working on HCA tickets now, so tagging to keep in the loop (I'm unable to assign as can not edit permissions).

Also tagging @dosumis.

Please note I'm out of office until 1st July, so I will get up to speed with any discussion on this ticket then.

[ami-day]

Hi @ami-day, just to update you that Paola is no longer working on HCA. @bvarner-ebi will be working on HCA tickets now, so tagging to keep in the loop (I'm unable to assign as can not edit permissions).

Also tagging @dosumis.

Please note I'm out of office until 1st July, so I will get up to speed with any discussion on this ticket then.

Hi @zoependlington thanks for letting me know! I'll assign @dosumis and @bvarner-ebi to the ticket, and correspond with them until you're back in the office.

[ami-day]

I don't seem to be able to find and assign @bvarner-ebi to the ticket. Is their Github handle not yet linked to the HumanCellAtlas Ontology organisation/group?

[ESapenaVentura]

I have invited @bvarner-ebi to the repo, once the invitation is accepted we should be able to tag him 👍

[dosumis]

Odd use of GO BP terms - these refer to endogenous processes, not experimental treatments. An EFO term may be more appropriate - in which case @zoependlington can add when she returns. May also be useful to have a protocols.io link?

[ESapenaVentura]

@dosumis - The GO terms we are requesting are for the target_pathways field, so I think it makes sense that we pick GO terms for that - For the method, we will actually stick to terms under "OBI:0000070" (Assay). Do you think that EFO would be better?

[dosumis]

@dosumis - The GO terms we are requesting are for the target_pathways field, so I think it makes sense that we pick GO terms for that - For the method, we will actually stick to terms under "OBI:0000070" (Assay).

OK. That makes sense.

Do you think that EFO would be better?

Let's discuss with @zoependlington when she's back. That OBI branch is pretty huge + It's not clear to me how responsive they are.

[ami-day]

Thank you for your suggestions! After discussion with @ESapenaVentura we have agreed that for this particular use case, the GO terms would be best (target_pathways field), especially as it is quite broad (and we expect target_pathway to vary quite a bit). This ontology is indicated in the relevant schema update.