NCBITaxon to CHEBI via location_of -- weird location and also predicate upgrade

[realmarcin]

We have two different CHEBI's for location of:

CHEBI:33290
CHEBI:24632

One of these is in the opposite orientation:

urn:uuid:787caa6b-dd79-479c-b730-cb19a11ac3b2 CHEBI:24632 biolink:location_of NCBITaxon:930171 RO:0001015 Graph
urn:uuid:2ca87a3f-ecdb-4f81-b0e3-b61d43e71928 NCBITaxon:1441714 biolink:location_of CHEBI:33290 RO:0001015 Graph

CHEBI:24632 is hydrocarbon -- but doesn't sound right as a location, so something is off here. I think this is from Madin et al. traits.csv but the closest term I see in the raw file is "hydrocarbon_degradation" but this is for a 'metabolism' and not location.

Actually all of isolation_source data (from BacDive) is in the wrong orientation semantically -- let's just change the predicate to 'located_in' and keep the taxon as the subject.

@bsantan @hrshdhgd

[hrshdhgd]

Both these are coming from environments.csv which is part of the ingest. Look at the last column ENVO_ids

CHEBI:33290: line 86
CHEBI:24632: line 15

[hrshdhgd]

Actually all of isolation_source data (from BacDive) is in the wrong orientation semantically -- let's just change the predicate to 'located_in' and keep the taxon as the subject.

I don't see either CURIEs in the Bacdive transform.