From 989b5a0044625833664083f5dc45fef8bf156490 Mon Sep 17 00:00:00 2001 From: Damien Goutte-Gattat Date: Thu, 26 Oct 2023 09:26:12 +0100 Subject: [PATCH] Fix composite unsats (#3099) * Remove old bogus axiom from the local CTENO mirror. The local CTENO mirror, used to produce composite-metazoan, contains several axioms coming from an old version of Uberon (CTENO has not been updated since 2016). One such axiom is 'surface groove' SubClassOf: 'surface structure' which *may* have been correct in 2016 but is clearly bogus now, since 'surface groove' (UBERON:0006846) is an immaterial anatomical entity whereas 'surface structure' (UBERON:0003102) is an anatomical structure (and thus a material anatomical entity). That single axiom is causing massive unsatisfiability issues when building composite-metazoan. * Remove bogus axiom from the local EHDAA2 mirror. The local EHDAA2 mirror, used to build composite-metazoan, contains the following axiom: 'extra embryonic coelum' SubClassOf: 'attached to' some 'coelomic cavity lumen' This is wrong as 'attached to' (RO:0002371) is range- and domain-restricted to anatomical structures, while both 'extra embryonic coelum' (EHDAA2:0004739) and 'coelomic cavity lumen' (EHDAA2:0000267) are anatomical spaces (immaterial entities). This should really be fixed in EHDAA2, but the ontology is officially marked as inactive and has not seen any really meaningful edit for the past 8 years. * Unmap 'serous sac' from EMAPA:16060 and MA:0000005. 'serous sac' (UBERON:0005906) is mapped with EMAPA:16060. This is wrong, because the Uberon term represents an entire organ, made of a serous membrane surrounding an anatomical cavity, while the EMAPA term refers to a body cavity OR the lining that surrounds it. The Uberon term is necessarily a material entity, while the EMAPA term is not committed to being either a material entity or an immaterial entity. Mapping the two terms together results, when building composite-metazoan, in the EMAPA term becoming a subclass of the Uberon term, thereby forcing the EMAPA term to be a material entity. This causes several unsatisfiability issues because there are other Uberon terms that are mapped to subclasses of the EMAPA term but represent immaterial entities. Likewise for MA:0000005. --- src/ontology/imports/local-cteno.owl | 6 ------ src/ontology/imports/local-ehdaa2.owl | 3 +-- src/ontology/uberon-edit.obo | 2 -- 3 files changed, 1 insertion(+), 10 deletions(-) diff --git a/src/ontology/imports/local-cteno.owl b/src/ontology/imports/local-cteno.owl index 6b50117eb0..d019a013ec 100644 --- a/src/ontology/imports/local-cteno.owl +++ b/src/ontology/imports/local-cteno.owl @@ -12857,12 +12857,6 @@ - - - - - surface groove - diff --git a/src/ontology/imports/local-ehdaa2.owl b/src/ontology/imports/local-ehdaa2.owl index 9400641d14..5afe4ae16e 100644 --- a/src/ontology/imports/local-ehdaa2.owl +++ b/src/ontology/imports/local-ehdaa2.owl @@ -25905,7 +25905,6 @@ SubClassOf( ObjectSomeValuesFrom( AnnotationAssertion(rdfs:label "extraembryonic coelum"^^xsd:string) SubClassOf( ) SubClassOf( ObjectSomeValuesFrom( )) -SubClassOf( ObjectSomeValuesFrom( )) SubClassOf( ObjectSomeValuesFrom( )) # Class: (urogenital sinus epithelium vesical part) @@ -26330,4 +26329,4 @@ AnnotationAssertion( "true"^^xsd:boolean) -) \ No newline at end of file +) diff --git a/src/ontology/uberon-edit.obo b/src/ontology/uberon-edit.obo index 9b6a1ff8a4..d806382e30 100644 --- a/src/ontology/uberon-edit.obo +++ b/src/ontology/uberon-edit.obo @@ -106778,9 +106778,7 @@ id: UBERON:0005906 name: serous sac def: "Organ with organ cavity, which has as parts a serous membrane and a serous cavity . Examples: pleural sac, pericardial sac, tendon sheath, bursa.[FMA]." [FMA:9689, https://github.com/obophenotype/uberon/issues/86] comment: See notes for serous membrane -xref: EMAPA:16060 xref: FMA:9689 -xref: MA:0000005 xref: SCTID:362890006 is_a: UBERON:0000062 ! organ relationship: has_part UBERON:0000042 ! serous membrane