Open cmungall opened 4 years ago
@cmungall we created for Rhea, so that they can use it. However ChEBI is not using it. @amalik01 will comment on the rest.
FYI, the predicate has_major_microspecies_at_pH_7_3 is populated in our local version of ChEBI (see ftp://ftp.expasy.org/databases/rhea/rdf/chebi.owl.gz) This chebi.owl file corresponds to data used in Rhea and UniProt. To be consistent ChEBI, Rhea, UniProt, ENZYME, etc are frozen then published in sync.
It recently came to my attention that CHEBI has a relationship type has_major_microspecies_at_pH_7_3, however this does not actually seem to be used in the OWL file:
http://www.ontobee.org/ontology/CHEBI?iri=http://purl.obolibrary.org/obo/chebi%23has_major_microspecies_at_pH_7_3
@amorgat had some context here: https://github.com/geneontology/go-ontology/issues/17763#issuecomment-674964438
in this thread became apparent that is-a hierarchies are not consistently populated between conjugate acids, bases, and zwitterions. Can this not be automated using OWL reasoning?
Anne also provided this very helpful set of slides on how Rhea does classification using CHEBI. This is very useful but seems overly complicated. Other ontologies in OBO avoid this kind of complexity by simply having a generic form for different sub-forms and having complete is-a relationships here. Is this not possible for CHEBI?