Rename COB 'molecular entity' to 'molecule'

[nataled]

As indicated in #104, COB's 'molecular entity' is much closer in spirit to CHEBI's 'molecule' than to CHEBI's 'molecular entity'. Accordingly, I propose that the current 'molecular entity' in COB be renamed to 'molecule'. It's clearer, less inclined toward misinterpretation, and better fits with the naming for other terms beneath 'material entity'.

[cmungall]

COB's definition covers functional groups, seems a little unintuitive to call these molecules which implies (IMO) wholeness/object/maximally connected (covalently) component?

[nataled]

(Definitions reproduced/updated here for ease of reference)

ChEBI molecular entity: Any constitutionally or isotopically distinct atom, molecule, ion, ion pair, radical, radical ion, complex, conformer etc., identifiable as a separately distinguishable entity.
COB molecular entity: A material entity that consists of two or more atoms that are all connected via covalent bonds such that any atom can be transitively connected with any other atom. This molecular entity is different than ChEBI's 'molecular entity'. We would like to have cardinality restrictions on the logic, but there are some technical limitations.
ChEBI molecule: Any polyatomic entity that is an electrically neutral entity consisting of more than one atom. (NOTE from DN: molecule is_a polyatomic entity is_a molecular entity)

@cmungall I see your point. For me the key aspect of the CHEBI 'molecular entity' definition is "identifiable as a separately distinguishable entity" (which, to my mind, also includes functional groups). More specifically--and key to how I would use these terms--CHEBI's definition allows the inclusion of complexes.

The concern for me is a practical one: There has been a proposal (memory now fuzzy on this) that all terms in OBO ontologies must correspond to a term in COB (this was not a requirement when COB was first proposed/discussed; it was proposed later). I actually like and agree with the proposal, but...

In PRO, we have what is, in essence, a new upper level term "protein-containing molecular entity" intended to cover individual proteins (and processed parts thereof), protein complexes, and protein aggregates. As things stand now, this term will be free standing, so to speak. That is, there is no COB term that corresponds to it. I was hoping that COB would replicate the spirit of CHEBI's 'molecule' and 'molecular entity'. Of course, if the proposal is not adopted, then PRO can happily stay under CHEBI's molecular entity instead of COB's (with, of course, the ensuing confusion that the identically-named but non-equivalent terms will generate).