Term issue: proposed obsoletion: protein-containing material entity

[cmungall]

I propose obsoleting the following term:

id: PR:000050567
name: protein-containing material entity
def: "A material entity that minimally consists of a protein." []
synonym: "protein" NARROW []
synonym: "protein aggregate" NARROW []
synonym: "protein complex" NARROW []
synonym: "protein-containing complex" NARROW []
relationship: has_part PR:000000001
is_a: BFO:0000040
intersection_of: BFO:0000040
intersection_of: BFO:0000051 PR:000000001

rationale:

• the extent of the class is outside the scope of PRO, encompassing organs, organisms, a chunk of soil, a building, the Atlantic ocean, the biosphere, the universe
• there is no documented use case for it
• the synonyms are incorrect (even when marked with NARROW).
• (main reason) this causes ragged incomplete lattices when PR is imported

The last (main) point deserves some explanation. When a module is extracted from PRO using SLME this class is typically included, and then after reasoning some classes may be classified here based on has-part relationships; however, this classification is always massively incomplete, as well as being flat and confusing

For example, in CL, there is a giant flat list of arbitrary terms under this:

https://www.ebi.ac.uk/ols4/ontologies/cl/classes/http%253A%252F%252Fpurl.obolibrary.org%252Fobo%252FPR_000050567

the arbitrariness can be seen if we focus on two terms, say, neuron and mast cell:

this is very confusing to a user, and although no ontology is ever complete (open world assumption), we should try to avoid creating grouping classes that lead to proliferation of incompleteness. See open world considered harmful.

[nataled]

You'll be glad to know we're actually way ahead of you on this, as I noticed the ill effect of the term on CL some time ago, and alerted Alex. It was intended to be protein-containing molecular entity. I'm keen to get rid of the offending term ASAP, but two things are in the way: a) Principle 13, which requires us to give advanced notice. At the moment our only avenue for doing this is via PRO release notes, which would mean it'll take two releases. b) Even worse, one of our key developers has left the project, delaying releases.

We're working diligently on b. Regarding a, I'm thinking to use OBO Discuss to make the announcement so it can be done with the next release. It's only been out for a single release (though that was back in February).

[cmungall]

Excellent! I think the timeline for (a) is acceptable, CL/UBERON can independently take action to exclude this from extracts (see linked issue).