obsolete should supersede inactive

[cmungall]

Following from

• 1721

There is a lovely rainbow of colors here:

https://obofoundry.org/ontology/ehda

I think obsolete implies inactive, we should either constrain the schema to allow either but not both, or we should filter at the UI level

The mixed-case indicator is a false positive but that is reported in another issue

Another oddity here is that the replacement for ehda (ehdaa2) is inactive (not not obsolete). I think this is OK and valid but it may be confusing to some. We may also want to make suggestions for active alternatives for ehdaa2 (this would be uberon or fma, the latter which now includes more embryonic terms but is still less complete than ehdaa2). Some of this may be best done in plain old fashioned text.

[matentzn]

• [X] inactive and orphaned banners should only show up when not deprecated (https://github.com/OBOFoundry/OBOFoundry.github.io/pull/1738).
• [X] irregular preferred prefix check only if preferred prefix is set (https://github.com/OBOFoundry/OBOFoundry.github.io/pull/1738).
• [ ] Should inactive ontologies be referable by the consider relation?

[nlharris]

did anyone make a decision about the last point above?

[matentzn]

I think no.

@cthoyt Do you know if there are ontologies being referred to by consider that are marked as inactive?

[cthoyt]

import requests
import yaml

URL = "https://raw.githubusercontent.com/OBOFoundry/OBOFoundry.github.io/master/registry/ontologies.yml"

def main():
    records = yaml.safe_load(requests.get(URL).text)["ontologies"]
    prefixes = {
        record["id"]
        for record in records
        if record["activity_status"] != "active"
    }
    for record in records:
        replaced_by = record.get("replaced_by")
        if replaced_by in prefixes:
            print("-", record["id"], "replaced by non-active ontology", replaced_by)

if __name__ == '__main__':
    main()

Gives back:

• ehda replaced by non-active ontology ehdaa2
• ehdaa replaced by non-active ontology ehdaa2

[matentzn]

If you would be so kind to add a test to disallow this and remove the consider relation from these two, that would be very appreciated :) This will close this ticket here as well.

[cmungall]

ehda and ehdaa were replaced by ehdaa2. we should not lose this information

[cthoyt]

i have no opinion one way or another. The text inside ehdaa2 about alternatives isn't super actionable, though - it currently reads:

This ontology has been inactive for several years, as of 2022. As of yet there is no complete replacement ontology, but OBO users may want to consider potential alternatives:

• HsapDv contains human-specific embryonic stage terms for Carnegie stages
• Uberon includes embryonic anatomy and developmental stage relations, but is more taxonomically general than EHDAA2, and may be less complete and less precise. Uberon includes EHDAA2 in its composite metazoan build.
• FMA now includes some himan embryonic anatomy, but is constructed on different principles than EHDAA2, and may be less complete and less precise in places

As of 2022, EHDAA2 still has the most complete set of human structure to stage relationships of any ontology in OBO, but note that it is no longer updated.

[cmungall]

That's an accurate reflection on the current situation! I could write more with specific guidance based on what the task is but I think it would just be better to direct people to make an issue on the uberon tracker and have a discussion

[matentzn]

Is replaced_by intended to capture provenance, or intended to guide the OBO Foundry users to our reference ontologies? If its just about provenance, then I totally misunderstood its purpose and we wouldn't constrain the field..

[nlharris]

I feel that the issue originally reported here has been fixed. Are there still action items in this ticket?

[nlharris]

Closing; please open a new ticket if there are still actionable tasks.