What relation should link a life stage term to its taxon-specific counterpart?

[gouttegd]

(This question was first raised in the developmental stage ontologies repo, but as a side remark on a PR. It deserves its own discussion.)

Part of the steps that are needed to build Composite Metazoan is to create “bridging axioms” that link the taxon-neutral terms in Uberon to their taxon-specific counterparts in the the taxon-specific ontologies (FBbt, FBdv, WBbt, XAO, etc.).

For almost all terms, that link is of the form:

{taxon-specific term} EquivalentTo: {Uberon term} and ('part of' some {taxon ID})

But for the terms that represent life stages (specifically 'life cycle' and 'life cycle stage' and their descendants), the links can use two distinct relations depending on where the bridging axioms are generated.

For the bridging axioms that are generated in the life stage ontologies repository, they are of the same form as above (in particular, they use the same part of relation).

But for the bridging axioms that are generated in Uberon (typically, those are the bridges that link to externally maintained, taxon-specific life stage ontologies like FBdv or WBls), the links are of the form

{taxon-specific term} EquivalentTo: {Uberon term} and ('occurs in' some {taxon ID})

where occurs in is BFO:0000066.

It cannot be good that the relations used in those bridging axioms vary depending on the species (e.g. occurs in for Drosophila melanogaster, part of for Gorilla gorilla). Whatever relation is used, it should be the same for all species.

[gouttegd]

@cmungall Two years ago you suggested that all links between Uberon terms and taxon-specific terms should take the form of two axioms:

{taxon-specific term} EquivalentTo: {Uberon term} and ('in taxon' some {taxon ID})
{taxon-specific term} SubClassOf: ({relation} some {taxon ID})

where relation would be 'part of' for continuants and 'occurs in' for occurrents. Is that still what you think?