Use KP weights for a benchmark

[cbizon]

KPs return weights for their edges of various types, but to what degree to ARAs make use of them?

Query COHD for disease - biolink:has_real_world_evidence_of_association_with - chemical where the disease is pinned, for example, to cystic fibrosis. COHD returns a multitude of metrics that could be used to rank results, create benchmarks for those metrics to determine how e.g. the ARAGORN rankings match the COHD rankings.

Example query:

{
  "workflow": [
    {
      "id": "fill"
    },
    {
      "id": "overlay_compute_ngd",
      "parameters": {
        "virtual_relation_label": "N1",
        "qnode_keys": [
          "N_disease",
          "N_known_chemical"
        ]
      }
    },
    {
      "id": "bind"
    },
    {
      "id": "complete_results"
    },
    {
      "id": "score"
    },
    {
      "id": "filter_results_top_n",
      "parameters": {
        "max_results": 100
      }
    }
  ],
  "message": {
        "query_graph": {
            "nodes": {
                "N_disease": {
                     "ids": ["MONDO:0009061"],
                     "categories": ["biolink:Disease"]
                },
                "N_known_chemical": {
                    "categories": ["biolink:ChemicalEntity"]
                }
            },
            "edges": {
                "E_dis_chem": {
                    "subject": "N_disease",
                    "object": "N_known_chemical",
                    "predicates": ["biolink:has_real_world_evidence_of_association_with"]
                }
            }
        }
    }
}