revert MIREOT style import of GO's "protein trimerization" in #36

[StroemPhi]

In #36 GO's "protein trimerization" was imported into MOP as a subclass of trimerization (MOP:0000715). This will most likely lead to issues downstream if MOP and GO are to be combined and protein trimerization being defined thus as a subclass of trimerization in MOP and also a subclass of protein complex oligomerization in GO.

[StroemPhi]

@colinbatchelor it might be better to resume the discussion about this axiom injection here, instead of my closed PR.
As said there, injecting this subclass of axiom on protein trimerization would probably be considered problematic. Especially, since we didn't tag it as an injected axiom or have a "base" MOP release, which doesn't contain it. Having said that, the question is, how we proceed.
I see two possible solutions:

1. drop protein trimerization from MOP but add an editor note which links to this issue, in order to avert the bad code smell of having cyclotrimerization as the only child of trimerization
2. consult with the GO and/or OBO/COB team where to assert this bridge axiom in order to be OBO principles compliant