TRAPI 1.4 validation error from automat-gtopdb edge: empty attribute value

[colleenXu]

I was doing a quick performance review of the BTE "chem-affects-gene" templates, and I used my notebook to run TRAPI validation on some responses.

The response that raised this validation error is > 25 MB so I can't attach it, but I'll include the query I ran below. Notes on how I ran the query:

• it took me 3 min 2 s to run
• had 500 results
• I was running with INSTANCE_ENV=ci (so the server urls for automat KPs were CI instance ones)

query: gabapentin increases activity/abundance of what?

{
    "message": {
        "query_graph": {
            "nodes": {
                "start": {
                    "categories": ["biolink:ChemicalEntity"],
                    "ids": ["PUBCHEM.COMPOUND:3446"]
                },
                "end": {
                    "categories": ["biolink:Gene"]
                }
            },
            "edges": {
                "t_edge": {
                    "subject": "start",
                    "object": "end",
                    "predicates": ["biolink:affects"],
                    "knowledge_type": "inferred",
                    "qualifier_constraints": [
                        {
                            "qualifier_set": [
                                {
                                    "qualifier_type_id": "biolink:object_aspect_qualifier",
                                    "qualifier_value": "activity_or_abundance"
                                },
                                {
                                    "qualifier_type_id": "biolink:object_direction_qualifier",
                                    "qualifier_value": "increased"
                                }
                            ]
                        }
                    ]
                }
            }
        }
    }
}

The reasoner-validator raised the error: error.knowledge_graph.edge.attribute.value.empty. Note that "errors" are "orange" in the ARAX UI, vs the red for "critical" errors.

This error comes from 1 specific KG edge, which is from Automat GtopDB. Example error code and object from my notebook (python):

error.knowledge_graph.edge.attribute.value.empty

{'infores:gtopdb -> infores:automat-gtopdb -> infores:biothings-explorer': {'PUBCHEM.COMPOUND:3446[biolink:SmallMolecule]--biolink:affects->NCBIGene:3786[biolink:Gene]': None}}

---

I'm not sure why this validation error is being raised. I think I've identified the edge...but its edge-attributes don't have empty stuff in the value field. Maybe it's the value false (which only shows up as an edge-attribute value in this edge)?

Edge ID: 3e21149474a2f5bb54ebe73cee7984dd

edge contents

{
  "predicate": "biolink:affects",
  "subject": "PUBCHEM.COMPOUND:3446",
  "object": "NCBIGene:3786",
  "qualifiers": [
    {
      "qualifier_type_id": "biolink:object_direction_qualifier",
      "qualifier_value": "increased"
    },
    {
      "qualifier_type_id": "biolink:qualified_predicate",
      "qualifier_value": "biolink:causes"
    },
    {
      "qualifier_type_id": "biolink:object_aspect_qualifier",
      "qualifier_value": "activity"
    }
  ],
  "attributes": [
    {
      "attribute_type_id": "aragorn:endogenous",
      "value": false,
      "value_type_id": "xsd:boolean",
      "original_attribute_name": "endogenous",
      "value_url": null,
      "attribute_source": null,
      "description": null,
      "attributes": null
    },
    {
      "attribute_type_id": "biolink:Attribute",
      "value": 8.279999732971191,
      "value_type_id": "EDAM:data_0006",
      "original_attribute_name": "affinity",
      "value_url": null,
      "attribute_source": null,
      "description": null,
      "attributes": null
    },
    {
      "attribute_type_id": "biolink:Attribute",
      "value": false,
      "value_type_id": "EDAM:data_0006",
      "original_attribute_name": "primaryTarget",
      "value_url": null,
      "attribute_source": null,
      "description": null,
      "attributes": null
    },
    {
      "attribute_type_id": "biolink:publications",
      "value": [
        "PMID:30021858"
      ],
      "value_type_id": "EDAM:data_0006",
      "original_attribute_name": "publications",
      "value_url": null,
      "attribute_source": null,
      "description": null,
      "attributes": null
    },
    {
      "attribute_type_id": "biolink:Attribute",
      "value": "pEC50",
      "value_type_id": "EDAM:data_0006",
      "original_attribute_name": "affinityParameter",
      "value_url": null,
      "attribute_source": null,
      "description": null,
      "attributes": null
    }
  ],
  "sources": [
    {
      "resource_id": "infores:automat-gtopdb",
      "resource_role": "aggregator_knowledge_source",
      "upstream_resource_ids": [
        "infores:gtopdb"
      ],
      "source_record_urls": null
    },
    {
      "resource_id": "infores:gtopdb",
      "resource_role": "primary_knowledge_source",
      "upstream_resource_ids": null,
      "source_record_urls": null
    },
    {
      "resource_id": "infores:biothings-explorer",
      "resource_role": "aggregator_knowledge_source",
      "upstream_resource_ids": [
        "infores:automat-gtopdb"
      ]
    }
  ]
}

[colleenXu]

I consider this less important and unlikely to cause issues downstream...but maybe it'd be good to ask about this anyways?

And I think this should be brought up with Automat devs (Ranking Agent?) and that they should take the lead on data-modeling / reasoner-validator discussions.

[andrewsu]

Per this discussion on slack, this error is a false positive by the validator, to be fixed in the next release. Closing this issue...

[colleenXu]

I used reasoner-validator 3.8.6 locally to validate the same saved response that originally triggered this error.

The error wasn't raised anymore.