Pickier ingest of xrefs/associated_with

[jsstevenson]

For source data that provides qualifiers/scope for xrefs, I think we'd like to stick with narrow mappings (eg exactMatch, equivalentTo) rather than broad mappings (eg relatedTo). There are cases where we aren't doing this, though, such as Mondo (example: in Mondo, Hirschsprung disease-ganglioneuroblastoma syndrome includes a relatedTo reference to Orphanet:635 (regular neuroblastoma) (link)

[jsstevenson]

Now that we're commiting to OBO in #171, some more context. Here's the example for MONDO:0013082, "Hirschsprung disease-ganglioneuroblastoma syndrome":

[Term]
id: MONDO:0013082
name: Hirschsprung disease-ganglioneuroblastoma syndrome
def: "A rare, genetic, developmental defect during embryogenesis syndrome characterized by total or partial colonic aganglionosis associated with peripheral, usually multifocal, neuroblastic tumors (ganglioneuroblastoma, neuroblastoma, ganglioneuroma). Congenital central hypoventilation syndrome, with variable severity of respiratory compromise, cardiovascular and ophthalmologic symptoms, consistent with autonomic nervous system dysfunction, is occasionally associated." [Orphanet:2151]
subset: ordo_malformation_syndrome {source="Orphanet:2151"}
subset: predisposition
synonym: "Hirschsprung disease ganglioneuroblastoma" RELATED [GARD:0002695]
synonym: "neuroblastoma with Hirschsprung disease" RELATED [OMIM:613013]
xref: MESH:C538119 {source="Orphanet:2151/e", source="MONDO:equivalentTo", source="Orphanet:2151"}
xref: Orphanet:2151 {source="MONDO:equivalentTo"}
xref: Orphanet:635 {source="OMIM:613013", source="MONDO:relatedTo"}
xref: UMLS:C2751683 {source="OMIM:613013", source="Orphanet:2151/e", source="MONDO:equivalentTo", source="Orphanet:2151"}
is_a: MONDO:0015246 {source="Orphanet:2151"} ! syndromic anorectal malformation
is_a: MONDO:0021189 {source="Orphanet:2151"} ! intestinal motility disease
property_value: exactMatch http://identifiers.org/mesh/C538119
property_value: exactMatch http://linkedlifedata.com/resource/umls/id/C2751683
property_value: exactMatch Orphanet:2151

Orphanet:635 is provided as a "related" concept in a trailing qualifier. Unfortunately, OBO parsers don't really have to include material in that section:

Parser implementations may choose to decode and/or round-trip these trailing modifiers. However, this is not required. A parser may choose to ignore or strip away trailing modifiers.

For this reason, trailing modifiers should only include information that is optional or experimental.

https://owlcollab.github.io/oboformat/doc/GO.format.obo-1_4.html#S.1.4

The fastobo Rust crate does parse them, but fastobo-py doesn't provide a way to capture them, yet.

Another option is to look into the SSSOM-provided mappings, because they probably can be used to scope xrefs.

[github-actions]

This issue is stale because it has been open 135 days with no activity. This issue will be closed if no further activity occurs in 14 days.