fix: warn+terminate if edge drops all results

[tokebe]

Addresses cases where sub-queries return results, but those results aren't kept because they don't match the opposing qNode ID(s).

After updating edge results, check if the edge still has any results and terminate the query if it doesn't, with a warning to the user via TRAPI logs.

[colleenXu]

Example queries that would get a 500 error with the current code, and are now correctly handled by this PR:

the tyrosine-phenylalanine query from the linked issue

POST to http://localhost:3000/v1/smartapi/1d288b3a3caf75d541ffaae3aab386c8/query (semmeddb thru BTE specifically)

{
  "message": {
    "query_graph": {
      "edges": {
        "e00": {
          "subject": "n00",
          "object": "n01",
          "predicates": [
            "biolink:derives_from"
          ]
        }
      },
      "nodes": {
        "n00": {
          "categories": [
            "biolink:Polypeptide"
          ],
          "ids": [
            "UMLS:C0041485"
          ]
        },
        "n01": {
          "categories": [
            "biolink:ChemicalEntity"
          ],
          "ids": [
            "UMLS:C0031453"
          ]
        }
      }
    }
  }
}

one-hop Explain related to demo query D3

{
  "message": {
    "query_graph": {
      "edges": {
        "e00": {
          "subject": "n00",
          "object": "n01"
        }
      },
      "nodes": {
        "n00": {
          "ids": ["DRUGBANK:DB00715"],
          "categories": ["biolink:SmallMolecule"]
        },
        "n01": {
          "ids": ["DOID:1936"],
          "categories": ["biolink:Disease"]
        }
      }
    }
  }
}

two-hop Explain related to demo query A2a

{
    "message": {
        "query_graph": {
            "edges": {
                "e01": {
                    "object": "n0",
                    "subject": "n1",
                    "predicates": [
                        "biolink:entity_negatively_regulates_entity"
                    ]
                },
                "e02": {
                    "object": "n1",
                    "subject": "n2",
                    "predicates":  [
                        "biolink:increases_abundance_of", 
                        "biolink:increases_expression_of", 
                        "biolink:increases_stability_of", 
                        "biolink:increases_uptake_of", 
                        "biolink:decreases_degradation_of", 
                        "biolink:increases_secretion_of", 
                        "biolink:increases_metabolic_processing_of", 
                        "biolink:increases_folding_of", 
                        "biolink:increases_localization_of", 
                        "biolink:increases_synthesis_of", 
                        "biolink:increases_response_to", 
                        "biolink:increases_splicing_of", 
                        "biolink:increases_mutation_rate_of", 
                        "biolink:increases_transport_of", 
                        "biolink:increases_activity_of", 
                        "biolink:increases_molecular_modification_of", 
                        "biolink:increases_molecular_interaction"
                    ]
                }
            },
            "nodes": {
                "n0": {
                    "ids": ["NCBIGene:23221"],
                    "categories": ["biolink:Gene"]
                },
                "n1": {
                    "categories": ["biolink:Gene"]
                },
                "n2": {
                    "ids": ["PUBCHEM.COMPOUND:15627"],
                    "categories": ["biolink:SmallMolecule"]
                }
            }
        }
    }
}

[colleenXu]

This looks to address the issue!

However, I'd want to test to see if there's any inadvertent changes to other queries (that are successful / have results).

[colleenXu]

I think of this issue as addressing the situation where two QNodes with IDs are connected to each other in the QGraph (like an Explain-style, but it can be more complex), but BTE doesn't actually find a connection between the two QNodes after all the sub-query execution and edge-management (intersecting records from different QEdges).

[colleenXu]

@tokebe do we need a test for this behavior?

[tokebe]

Yes, we could stand to test for this behavior.