Add instructions on how to set up an ubergraph instance with a custom set of ontologies

We would like an ubergraph instance with ENVO, PO, perhaps FOODON - and maybe also the LinkML OWL translations of MIxS (@hrshdhgd we will make this correspond with our OLS instance)

Should we set up our own instance and run on AWS? Or can we put into the main instance?

If we set up our own instance it looks like we should be able to figure from the Makefile, we can contribute back docs later

cc @turbomam @wdduncan @hrshdhgd

Start by reading the Makefile to identify how an ubergraph instance and custom ontology set are configured. Determine whether the documented path uses the main instance or a separate AWS deployment, then document the chosen setup and the ENVO, PO, FOODON, and possible MIxS inputs. Done means a newcomer can follow the instructions to create the requested instance.

Written by the indexing model from the issue text.