oio:hasAlternativeId should always has a referent, and it should always be a CURIE for a

[cmungall]

For context: owlcs/owlapi#317 -- a lot of this is my fault, I will endeavour to fix this. It only affects ontologies that implement GO-style merges (which @balhoff hates).

Identified by @althonos in EnvironmentOntology/envo#1045

The value of oio:hasAlternativeId

• must be a CURIE
• should always expand to a URI which
  • is not dangling [warn]
  • has owl:deprecated
  • has an annotation (http://purl.obolibrary.org/obo/IAO_0000231 http://purl.obolibrary.org/obo/IAO_0000227)
  • has no other axioms other than the above 2

robot report seems the appropriate place to fix this

here is a sample valid ontology:

############################
#   Annotation Properties
############################

# Annotation Property: <http://www.geneontology.org/formats/oboInOwl#hasAlternativeId> (has_alternative_id)

AnnotationAssertion(rdfs:label <http://www.geneontology.org/formats/oboInOwl#hasAlternativeId> "has_alternative_id"^^xsd:string)



############################
#   Classes
############################

# Class: <http://purl.obolibrary.org/obo/X_1> (<http://purl.obolibrary.org/obo/X_1>)

AnnotationAssertion(<http://purl.obolibrary.org/obo/IAO_0000231> <http://purl.obolibrary.org/obo/X_1> <http://purl.obolibrary.org/obo/IAO_0000227>)
AnnotationAssertion(<http://purl.obolibrary.org/obo/IAO_0100001> <http://purl.obolibrary.org/obo/X_1> <http://purl.obolibrary.org/obo/X_2>)
AnnotationAssertion(owl:deprecated <http://purl.obolibrary.org/obo/X_1> "true"^^xsd:boolean)

# Class: <http://purl.obolibrary.org/obo/X_2> (foo)

AnnotationAssertion(<http://www.geneontology.org/formats/oboInOwl#hasAlternativeId> <http://purl.obolibrary.org/obo/X_2> "X:1"^^xsd:string)
AnnotationAssertion(rdfs:label <http://purl.obolibrary.org/obo/X_2> "foo"^^xsd:string)
)

here is a sample invalid ontology

############################
#   Annotation Properties
############################

# Annotation Property: <http://www.geneontology.org/formats/oboInOwl#hasAlternativeId> (has_alternative_id)

AnnotationAssertion(rdfs:label <http://www.geneontology.org/formats/oboInOwl#hasAlternativeId> "has_alternative_id"^^xsd:string)



############################
#   Classes
############################

# Class: <http://purl.obolibrary.org/obo/X_1> (foo1)

AnnotationAssertion(<http://purl.obolibrary.org/obo/IAO_0000231> <http://purl.obolibrary.org/obo/X_1> <http://purl.obolibrary.org/obo/IAO_0000227>)
AnnotationAssertion(<http://purl.obolibrary.org/obo/IAO_0100001> <http://purl.obolibrary.org/obo/X_1> <http://purl.obolibrary.org/obo/X_2>)
AnnotationAssertion(rdfs:label <http://purl.obolibrary.org/obo/X_1> "foo1"^^xsd:string)
AnnotationAssertion(owl:deprecated <http://purl.obolibrary.org/obo/X_1> "true"^^xsd:boolean)

# Class: <http://purl.obolibrary.org/obo/X_2> (foo2)

AnnotationAssertion(<http://www.geneontology.org/formats/oboInOwl#hasAlternativeId> <http://purl.obolibrary.org/obo/X_2> "X:1"^^xsd:string)
AnnotationAssertion(rdfs:label <http://purl.obolibrary.org/obo/X_2> "foo2"^^xsd:string)

)

[althonos]

I'm confused, since I've always seen cases of alt_id being used to give an old identifier to a new term (typically, because of a merge), but it's probably because I'm not clear on the semantics. I understand the alt_id is a remnant of several terms being merged into one, but with what you propose you have the identifier referring to both the old and the new term?

I think I would have expected the replaced_by annotation property to be used here rather than the alt_id...