RTXteam
Unexpected use of qualifiers
While reviewing some ARAGORN creative mode queries, I ran across this edge:
{
"subject": "PUBCHEM.COMPOUND:5917",
"object": "MONDO:0005027",
"predicate": "biolink:regulates",
"qualifiers": [
{
"qualifier_type_id": "biolink:object_direction_qualifier",
"qualifier_value": "increased"
},
{
"qualifier_type_id": "biolink:object_aspect_qualifier",
"qualifier_value": "activity_or_abundance"
},
{
"qualifier_type_id": "biolink:qualified_predicate",
"qualifier_value": "biolink:causes"
}
],
"attributes": [
{
"attribute_type_id": "biolink:knowledge_source",
"value": "infores:semmeddb",
"value_type_id": "biolink:InformationResource",
"original_attribute_name": null,
"value_url": null,
"attribute_source": "infores:rtx-kg2",
"description": null,
"attributes": null
},
{
"attribute_type_id": "biolink:original_edge_information",
"value": [
"UMLS:C0030903---SEMMEDDB:augments---None---None---increased---UMLS:C0270844---SEMMEDDB:",
"UMLS:C0947690---SEMMEDDB:augments---None---None---increased---UMLS:C0036572---SEMMEDDB:",
"UMLS:C0030903---SEMMEDDB:augments---None---None---increased---UMLS:C0014544---SEMMEDDB:",
"UMLS:C0030903---SEMMEDDB:augments---None---None---increased---UMLS:C0234535---SEMMEDDB:",
"UMLS:C0947691---SEMMEDDB:augments---None---None---increased---UMLS:C0036572---SEMMEDDB:",
"UMLS:C0030903---SEMMEDDB:augments---None---None---increased---UMLS:C0751494---SEMMEDDB:",
"UMLS:C0030903---SEMMEDDB:augments---None---None---increased---UMLS:C0036572---SEMMEDDB:"
],
"value_type_id": "metatype:String",
"original_attribute_name": null,
"value_url": null,
"attribute_source": null,
"description": "The original RTX-KG2pre edge(s) corresponding to this edge prior to any synonymization or remapping. Listed in (subject)--(relation)--(object)--(source) format.",
"attributes": null
},
{
"attribute_type_id": "biolink:aggregator_knowledge_source",
"value": "infores:rtx-kg2",
"value_type_id": "biolink:InformationResource",
"original_attribute_name": null,
"value_url": null,
"attribute_source": "infores:rtx-kg2",
"description": null,
"attributes": null
},
{
"attribute_type_id": "biolink:aggregator_knowledge_source",
"value": "infores:aragorn",
"value_type_id": "biolink:InformationResource",
"original_attribute_name": null,
"value_url": null,
"attribute_source": "infores:aragorn",
"description": null,
"attributes": null
},
{
"value_type_id": null,
"original_attribute_name": null,
"value_url": null,
"attribute_source": "infores:semmeddb",
"description": null,
"attributes": null
},
...edited for length...
{
"original_attribute_name": "weight",
"attribute_type_id": "biolink:has_numeric_value",
"value": 1,
"value_type_id": "EDAM:data_1669",
"value_url": null,
"attribute_source": null
}
]
}
I'm surprised to find qualifiers here between a chemical and a disease - my understanding was that to date we only had qualified chem-gene edges (and I guess a gene-gene one).
Tagging for info: @sierra-moxon
Yes: we should only be implementing subject and object qualifiers between chemicals and genes. "abundance" in the qualifier enumeration: GeneOrGeneProductOrChemicalEntityAspectEnum, specifically refers to the abundance of some chemical or some gene product, rather than an observation of the severity(?) or population abundance(?) of disease.
We can certainly tackle gene->disease/phenotype qualifier use cases with priority on the modeling side. Can you give me an idea of the full statement that we are trying to make here? Is it about severity of disease or population abundance or something else?
