suppressor T lymphocyte in FMA is mapped to multiple CL classes

[cmungall]

...causing this in uberon

We need to get these checks in CL release too. We have treat-xrefs-as-reverse-genus-differentia: FMA part_of NCBITaxon:9606, i.e. FMA:x = CL:y and part_of some Homo sapiens

This means the xrefs have to be 1:1

$ grep FMA:84070 cl-edit.owl 
AnnotationAssertion(<http://www.geneontology.org/formats/oboInOwl#hasDbXref> <http://purl.obolibrary.org/obo/CL_0000792> "FMA:84070"^^xsd:string)
AnnotationAssertion(<http://www.geneontology.org/formats/oboInOwl#hasDbXref> <http://purl.obolibrary.org/obo/CL_0000795> "FMA:84070"^^xsd:string)
AnnotationAssertion(<http://www.geneontology.org/formats/oboInOwl#hasDbXref> <http://purl.obolibrary.org/obo/CL_0000815> "FMA:84070"^^xsd:string)

[cmungall]

It's mapped to these 3:

given these all have a synonym of "suppressor T lymphocyte", scope BROAD, it makes sense only to retain the mapping at the level of the parent