loose classes? - Githubissues

mcourtot commented 6 years ago

https://www.ebi.ac.uk/ols/ontologies/mondo/

a number of MONDO IDs at the top level which are lacking any annotations and which PURLs don't resolve on Ontobee e.g.: http://purl.obolibrary.org/obo/MONDO_0009436

    <owl:Class rdf:about="http://purl.obolibrary.org/obo/Orphanet_2113">
        <obo:IAO_0100001 rdf:resource="http://purl.obolibrary.org/obo/MONDO_0009436"/>

where obo:IAO_0100001 is term replaced by

It seems like all those "loose classes" are stemming from the same type of declarations in the file.

cmungall commented 6 years ago

Ontobee is behind OLS, these were introduced in the latest release.

OK, here's what is happening. The obof is a bit easier to look at. This is the source file:

[Term]
id: MONDO:0009436
name: hypothalamic hamartomas
alt_id: Orphanet:2113 {source="OMIM:241800"}
comment: Editor note: obsoleted in ORDO
synonym: "Congenital Hypothalamic Hamartoma Syndrome" RELATED [OMIM:241800]
synonym: "HYPOTHALAMIC HAMARTOMAS" RELATED [OMIM:241800]
xref: MESH:C537158 {source="MONDO:ontobio", source="MONDO:equivalentTo"}
xref: OMIM:241800 {source="MONDO:Leader", source="MONDO:equivalentTo"}
xref: UMLS:C0342418 {source="OMIM:241800"}
is_a: MONDO:0007804 ! Pallister-hall syndrome

The ordo class is obsoleted:

https://www.ebi.ac.uk/ols/ontologies/ordo/terms?iri=http%3A%2F%2Fwww.orpha.net%2FORDO%2FOrphanet_2113

reason for obsolescence
This class is deprecated. The preferred class is use http://www.orpha.net/ORDO/Orphanet_672 with label: Pallister-Hall syndrome

however, in mondo we still want to track an explicit class for OMIM:241800. So we retain the ordo URI and have a replaced-by to the mondo node.

Now for releases, we map the mondo id onto its omim equivalent. We use the OWLAPI entity renamer. It seems maybe this doesn't work for things that are targets of annotation assertions?

Annoying issue will think of best approach

nicolevasilevsky commented 6 years ago

This issue was moved to monarch-initiative/mondo#156

monarch-initiative / monarch-disease-ontology-RETIRED

loose classes? #411