do we really need a chemical_entity_set?

[turbomam]

If we can find one or more namespaces that unambiguously identify reagents and metabolites, with high fidelity regarding the forms of those chemicals, then we don't need a chemical_entity_set. We just need slots that mention the external identifiers. Conventionally, those would be web-resolvable CURIes.

[kheal]

@kheal Tagging myself so I get updates.