Closed CaseyTa closed 2 years ago
@CaseyTa, indeed we have an issue with our metaKG - MolePro will return results for the following query, yet it does not include Gene-directly_interacts_with-SmallMolecule metaKG entry. We are going to fix that.
{
"message": {
"query_graph": {
"nodes": {
"n2": {
"ids":["NCBIGene:819"],
"categories": [
"biolink:Gene"
]
},
"n3": {
"categories": [
"biolink:SmallMolecule"
]
}
},
"edges": {
"e03": {
"subject": "n2",
"object": "n3",
"predicates": [
"biolink:directly_interacts_with"
]
}
}
}
}
}
MolePro's metaKG now declares both
"object": "biolink:Gene",
"predicate": "biolink:directly_interacts_with",
"subject": "biolink:SmallMolecule"
and
"object": "biolink:SmallMolecule",
"predicate": "biolink:directly_interacts_with",
"subject": "biolink:Gene"
Thanks, @vdancik!
MolePro is a potential source of information for the 3rd hop of Workflow B (query below). However, ARAX's expansion results UI says that MolePro doesn't support this edge: https://arax.ncats.io/?r=28066.
MolePro has the following in its MetaKG:
biolink:directly_interacts_with
is symmetric, but the inverse edge is not present in MolePro's MetaKG. I'm not sure if this should potentially be resolved on the ARAX or MolePro side. @edeutsch @vdancikNote, this issue ignores that ARAX is not calling KPs when there are more than 100 input CURIEs in the expansion.
Query: