Obsolete and merge a handful of redundant terms #3196
Conversation
The terms 'optic eminence surface ectoderm' and 'optic eminence ectoderm' appear to refer to the same concept, so we obsolete the former and merge it into the latter. closes #3156
The terms 'vitelline venous plexus' and 'venous vitelline plexus' are completely identifical except for the order of adjectives in the label. We obsolete the former and merge it into the latter. closes #3157
The term 'thymus subunit' is redundant with 'thymus lobe', so we obsolete it and replace it with 'thymus lobe'. closes #3158
The terms 'gland of anal sac' and 'gland of anal canal' seem to refer to the same thing. They both cross-reference the same Wikipedia page, the same NCIT and UMLS terms, and have both 'anal gland' as related synonym. We obsolete the former and merge it into the latter. closes #3159
The term 'telencephalic tracts' was created as a generalised version of the ZFA term 'telencephalic white matter' (defined as "CNS white matter that is part of the telencephalon"). It seems redundant with 'white matter of telencephalon' (defined as "a portion of white matter that is part of a telencephalon"). We obsolete 'telencephalic tracts' and merge it with 'white matter of telencephalon'. closes #3160
The term 'common crus' was created as a generalised version of the eponym ZFA term (defined as the "duct that forms when the nonampullary ends of the posterior and anterior semicircular canals merge"). It seems redundant with 'common crus of semicircular ducts' (defined as "a nonampullary end of the superior or posterior semicircular ducts"). There does not seem to be any other "common crus" than those of the semicircular ducts. We obsolete 'common crus' and merge it into 'common crus of semicircular ducts'. closes #3161
There was a problem hiding this comment.
Please, see inline comments.
The instructions in obo academy on obsoleting terms do not ask to introduce URL in replaced_by as IRI, I need to edit it and create a ticket to change all incorrect of obsoleted terms in Uberon, similar to CL.
| relationship: part_of UBERON:0005234 {is_inferred="true", source="https://github.com/obophenotype/uberon/wiki/Inferring-part-of-relationships"} ! optic eminence ectoderm | ||
| name: obsolete optic eminence surface ectoderm | ||
| comment: Obsoleted because it is redundant with 'optic eminence ectoderm'. | ||
| property_value: term_tracker_item "https://github.com/obophenotype/uberon/issues/3156" xsd:anyURI |
There was a problem hiding this comment.
Please, change the format of this URL to IRI
There was a problem hiding this comment.
Doing so results in the annotation being interpreted as a relationship, which violates one of the CI checks that tries to enforce that an obsolete term should have no relationships.
Who decided that the value of term_tracker_item should be an IRI anyway, and why was the tech group not consulted? This is going to be an issue for any ontology that uses the OBO format as its edit format.
There was a problem hiding this comment.
Who decided that the value of term_tracker_item should be an IRI anyway, and why was the tech group not consulted? This is going to be an issue for any ontology that uses the OBO format as its edit format.
Hi @anitacaron, could you provide more background on this, please?
There was a problem hiding this comment.
Correction: This may not be an issue for “any ontology that uses the OBO format” – merely for those that do the same kind of CI checks as Uberon.
There was a problem hiding this comment.
Technically, we want to ensure only URLs are added in term_tracker_item. But I agree that the issue is in the OBO Format.
There was a problem hiding this comment.
And what is the problem with storing URLs as literal URLs (xsd:anyURI)? Why insisting that they should be IRIs?
There was a problem hiding this comment.
In any case, either “we” drop this requirement that term_tracker_item should be IRIs, or “we” need to either:
- update our CI checks to make an exception that obsolete terms can have
term_tracker_itemrelationships, - update the OBO parser in the OWL API to stop interpreting IRI-valued annotations as relationships.
There was a problem hiding this comment.
All of the existing values for term_tracker_item in Uberon seem to be xsd:anyURI.
The “standard” for IAO:0000233 annotations is reportedly for the values to be represented as IRIs rather than literal URIs.
Term UBERON:0004760 is labelled as "gland of anal canal", so it does not need an exact synonym of "gland of anal canal" -- this is probably a remnant of a time where the term had another label.
|
Apologies for asking to change the format, it was a mistake from my part. Could you revert the commit, so this PR can be merged before the next release? |
|
I’ve pushed a new clean PR (same changes but without the 8d21d5f commit) to replace this one. |
This PR obsoletes a few terms that are redundant with other terms: