NCATSTranslator / testing

Materials and tools for testing Translator components
1 stars 9 forks source link

Gene (HGNC:11998) - is sequence variant of - SequenceVariant #108

Open sstemann opened 3 years ago

sstemann commented 3 years ago

Query: variant.json PK: 952b8413-09e9-4b52-8c62-1b5fd6ea6e3e Results Tracking Sheet

image

dkoslicki commented 3 years ago

Don’t know if this is due to a TRAPI issue or something with the ARS, but it appears the following works fine on ARAX:

{
  "edges": {
    "e00": {
      "exclude": false,
      "object": "n0",
      "predicates": [
        "biolink:is_sequence_variant_of"
      ],
      "subject": "n1"
    }
  },
  "nodes": {
    "n0": {
      "ids": [
        "HGNC:11998"
      ],
      "is_set": false
    },
    "n1": {
      "is_set": false
    }
  }
}

Results at https://arax.ncats.io/?r=23281

I tried submitting that same JSON to the ARS, but am getting a 404 for the ARS.

dkoslicki commented 3 years ago

My mistake, it appears to be due to lack of biolink:SequenceVariant nodes. If you use NamedThing or NucleicAcidEntity or similar subject node category, then more results are returned.

marcdubybroad commented 3 years ago

The genetics KP does not carry individual variants in our service, since we aggregate our variant statistic at the gene level for gene association numbers. So not getting any results from the genetics KP is expected. Please let us know if having variants exposed is of interest to the translator effort.

andrewsu commented 3 years ago

In the SmartAPI metadata for myvariant.info, we annotate the SequenceVariant - is_sequence_variant_of - Gene edge (see Block 1 below), but not (currently) the reverse Gene - has_sequence_variant - SequenceVariant. So that means that for this query (which specifies the gene ID as input) BTE does not find any matching APIs in the metaKG. If, however, we reverse the information by specifying a SequenceVariant (e.g., DBSNP:rs1555524108) and ask for the related gene (Block 2), BTE successfully returns with the correct answer.

This issue could be solved by adding the "reverse" operation to the myvariant.info SmartAPI metadata.

Block 1:

    variantAssociatedWithGene:
    - supportBatch: true
      inputSeparator: ","
      parameters:
        fields: dbsnp.gene.geneid
      predicate: is_sequence_variant_of
      source: "infores:dbsnp"
      requestBody:
        body:
          q: "{inputs[0]}"
          scopes: dbsnp.rsid
        header: application/x-www-form-urlencoded
      inputs:
      - id: DBSNP
        semantic: SequenceVariant
      outputs:
      - id: NCBIGene
        semantic: Gene
      response_mapping:
        "$ref": "#/components/x-bte-response-mapping/variant-gene"

Block 2:

{
    "message": {
        "query_graph": {
            "nodes": {
                "n0": {
                    "categories": [
                        "biolink:Gene"
                    ]
                },
                "n1": {
                    "categories": [
                        "biolink:SequenceVariant"
                    ],
                    "ids": "DBSNP:rs1555524108"
                }
            },
            "edges": {
                "e0": {
                    "subject": "n1",
                    "object": "n0",
                    "predicates": [
                        "biolink:is_sequence_variant_of"
                    ]
                }
            }
        }
    }
}
GregHydeDartmouth commented 3 years ago

CHP does not have information on biolink:SequenceVariant. 0 results is expected.