Disease (MONDO:0010161 Tyrosinemia type 1) - related to - Gene - related to Chemical Substance

[sstemann]

Query: tyroCycle.json PK: 588d1339-1d9e-4684-a62d-d18a712aa887 MONDO:0010161

Results Tracking Sheet

[brettasmi]

imProving Agent doesn't support cyclical queries at this time. Here is the ARAX dev info:

{
  "description": "imProving Agent currently only supports linear queries; returning empty message...",
  "message": {},
  "status": "Query unprocessable",
  "validation_result": {
    "message": "",
    "size": "0.51 kB",
    "status": "PASS",
    "version": "1.1.1"
  }
}

[dkoslicki]

ARAX’s query graph interpreter complains on cyclic queries (by default, it treats edges as undirected). However, the DSL (ARAXi) successfully executes, but finds no results

add_qnode(ids=MONDO:0010161, key=n0)
add_qnode(categories=biolink:Gene, key=n1)
add_qnode(categories=biolink:ChemicalSubstance, key=n2)
add_qedge(subject=n1, object=n0, key=e0)
add_qedge(subject=n1, object=n2, key=e1)
add_qedge(subject=n0, object=n2, key=e2)
expand(kp=RTX-KG2)
overlay(action=compute_ngd, virtual_relation_label=N1, subject_qnode_key=n0, object_qnode_key=n2)
overlay(action=compute_ngd, virtual_relation_label=N2, subject_qnode_key=n1, object_qnode_key=n2)
overlay(action=compute_ngd, virtual_relation_label=N3, subject_qnode_key=n1, object_qnode_key=n0)
resultify()
filter_results(action=limit_number_of_results, max_results=30)

So it appears to be that (at least in the RTX-KG2 KP), there are no paths found satisfying this query. If you remove edges e1 or e2, you do get results, but not when you require all of them. You can make one of the edges optional via:

add_qnode(ids=MONDO:0010161, key=n0)
add_qnode(categories=biolink:Gene, key=n1)
add_qnode(categories=biolink:ChemicalSubstance, key=n2)
add_qedge(subject=n1, object=n0, key=e0)
add_qedge(subject=n1, object=n2, key=e1)
add_qedge(subject=n0, object=n2, key=e2,option_group_id=b)
expand(kp=RTX-KG2)
overlay(action=compute_ngd, virtual_relation_label=N1, subject_qnode_key=n0, object_qnode_key=n2)
overlay(action=compute_ngd, virtual_relation_label=N2, subject_qnode_key=n1, object_qnode_key=n2)
overlay(action=compute_ngd, virtual_relation_label=N3, subject_qnode_key=n1, object_qnode_key=n0)
resultify()
filter_results(action=limit_number_of_results, max_results=30)

https://arax.ncats.io/?r=15960

Or

add_qnode(ids=MONDO:0010161, key=n0)
add_qnode(categories=biolink:Gene, key=n1)
add_qnode(categories=biolink:ChemicalSubstance, key=n2)
add_qedge(subject=n1, object=n0, key=e0)
add_qedge(subject=n1, object=n2, key=e1 ,option_group_id=a)
add_qedge(subject=n0, object=n2, key=e2)
expand(kp=RTX-KG2)
overlay(action=compute_ngd, virtual_relation_label=N1, subject_qnode_key=n0, object_qnode_key=n2)
overlay(action=compute_ngd, virtual_relation_label=N2, subject_qnode_key=n1, object_qnode_key=n2)
overlay(action=compute_ngd, virtual_relation_label=N3, subject_qnode_key=n1, object_qnode_key=n0)
resultify()
filter_results(action=limit_number_of_results, max_results=30)

https://arax.ncats.io/?r=15961

[dkoslicki]

The last URL is probably the intended results: https://arax.ncats.io/?r=15961

[rajshruti18]

CAM-KP does not have data on diseases so 0 results are expected. We do have data for edges between genes and chemical substance.