Use RHEA or EC to populate definitions

[pgaudet]

from #29562 (comment)

@cmungall

I think in this case the original GO def was just outright wrong and despite the provenance was neither sourced from EC or RHEA. I think this must just have been a copy and paste typo.

But here is what I propose moving forward:

every catalytic activity has a single source of truth (SoT); i.e ONE def_xref to EITHER EC or RHEA
exceptions should only be care cases where neither EC nor RHEA concept is appropriate and we make some kind of synthesis, and GO editors are the SoT. These should be rare, if they happen at all.
We automatically sync all definitions where there is a single SoT. We don't try and write complex string matching (e.g for + vs (+) and the many other chemical lexicogoraphy edge cases), and we don't use LLMs. We just drop and reload