add Association for has_substrate data

[colleenXu]

Urgent request for the Feb ingest work.

Chembl and DrugCentral have chem-protein substrate data, which is modeled as "Protein has_substrate Chemical". I couldn't find an existing Association that would be useful for validation.

So I'm proposing this new Association as a solution. I've made the subject/object categories broader to cover the various kinds of enzymes and substrates (in the hopes that it'll be a useful Association in future cases).

I've tagged @vdancik for his expertise and opinion on whether this works for the Chembl ingest. Currently, the ingest is usingGeneAffectsChemicalAssociation but this is problematic because has_substrate is NOT a descendant of affects...

---

Note:

I haven't tested this yet, because I'm still working on my DrugCentral ingest of this data (bioactivity table act_table_full)...

[vdancik]

@colleenXu , even though I am fine with using GeneAffectsChemicalAssociation, I see your point. If we go ahead with macromolecular machine has substrate association, we should also add macromolecular machine has product association - it will be needed when we start ingesting metabolic reactions.

[colleenXu]

@vdancik I'm surprised that there isn't a pydantic validation or pipeline validation error with GeneAffectsChemicalAssociation and the mismatched predicate...

Also, if this Association looks okay to you, can you approve this PR?