Synchronize with semsql build (i.e., do it here)

[caufieldjh]

Calling OAK on phenio is fastest with something like sqlite:obo:phenio - that retrieves the sqlite version of phenio from here: https://github.com/INCATools/semantic-sql
(or specifically, from here: https://s3.amazonaws.com/bbop-sqlite/)
That bucket won't always contain a sqlite representation of the most recent phenio, however, so it may not align with what's in kg-phenio.
If the sqlite representation is produced as part of the kg-phenio build process instead, then it should be in sync.