Clarify/fix is_substrate_of predicate definition

[mbrush]

See https://github.com/biolink/biolink-model/blob/master/biolink-model.yaml#L4623

is substrate of: is_a: participates in inverse: has substrate domain: chemical entity or gene or gene product range: chemical entity or gene or gene product

The is_substrate_of predicate is declared a child of participates in, which holds between an occurrent and a process, yet it is defined to hold between two continuants (chemical entities or genes or gene products). This is a contradiction.

The definition of participates_in needs work as well, per #1606. And then here, we need to decide if is_substrate_of describes the relationship between the continuant/chemical and the process/reaction in which it participates, or the continuant/chemical and the enzyme/catalyst that it is a substrate for.