Gene "negatively regulates" RHOBTB2 (NCBIGene:23221)

sstemann commented 3 years ago

Updated with the current predicate based on feedback from David Koslicki Query: negativelyRegulates.json PK: 33732257-d60c-40a3-a02a-33e0e7821799 NCBIGene: 23221 Control: Gene E2F1 (NCBIGene:1869) (BTE Result #181) Results Tracking Sheet

Results Returned From BTE ARAX (but not the control)

dkoslicki commented 3 years ago

The ARAX endpoint is currently using an apparently older version of the Biolink model, so the following query does return a result:

{
    "message": {
        "query_graph": {
            "edges": {
                "e01": {
                    "object": "n0",
                    "subject": "n1",
                    "predicates":["biolink:negatively_regulates_entity_to_entity"]
                }
            },
            "nodes": {
                "n0": {
                    "ids":["NCBIGene:23221"],
                    "categories": [
                        "biolink:Gene"
                    ]
                },
                "n1": {
                    "categories": [
                        "biolink:Gene"
                    ]
                }
            }
        }
    }
}

Note that here, Biolink uses negatively_regulates_entity_to_entity while here it uses entity_negatively_regulates_entity.

marcdubybroad commented 3 years ago

The genetics KP doesn't have gene/gene relationships, so no results expected.

bill-baumgartner commented 3 years ago

Text Mining Provider will have gene-regulates-gene assertions eventually (https://github.com/NCATSTranslator/Text-Mining-Provider-Roadmap/issues/7), but those assertions are not yet available.

dkoslicki commented 3 years ago

@andrewsu I find it curious that the ARAX results are not a superset of the BTE results since ARAX does call BTE for this query. We are sending BTE the following triples: biolink:Gene--biolink:negatively_regulates_entity_to_entity--(biolink:Gene, NCBIGene:23221) biolink:Gene--biolink:negatively_regulates_entity_to_entity--(biolink:Protein, NCBIGene:23221) biolink:Protein--biolink:negatively_regulates_entity_to_entity--(biolink:Protein, NCBIGene:23221) biolink:Protein--biolink:negatively_regulates_entity_to_entity--(biolink:Gene, NCBIGene:23221) But ARAX is getting no results from BTE. Any idea why this might be?

brettasmi commented 3 years ago

We ran into two issues here that have now been fixed. We're using the biolink master branch, which uses the predicate that @dkoslicki mentioned above. I re-ran the query using that predicate here: 4a2c270e-799a-4792-bca8-2c2197e3bdd6

pg427 commented 3 years ago

Explanatory Agent is currently not supporting the negatively_regulates_entity_to_entity predicate.

vdancik commented 3 years ago

No MolePro results expected since we do not have gene-regulates-gene information.

andrewsu commented 3 years ago

@andrewsu I find it curious that the ARAX results are not a superset of the BTE results since ARAX does call BTE for this query. We are sending BTE the following triples: biolink:Gene--biolink:negatively_regulates_entity_to_entity--(biolink:Gene, NCBIGene:23221) biolink:Gene--biolink:negatively_regulates_entity_to_entity--(biolink:Protein, NCBIGene:23221) biolink:Protein--biolink:negatively_regulates_entity_to_entity--(biolink:Protein, NCBIGene:23221) biolink:Protein--biolink:negatively_regulates_entity_to_entity--(biolink:Gene, NCBIGene:23221) But ARAX is getting no results from BTE. Any idea why this might be?

@dkoslicki For the TRAPI query you posted in https://github.com/NCATSTranslator/testing/issues/68#issuecomment-850050639, BTE returns a non-zero result. curl command below -- are you seeing something different?

curl --silent --location --request POST "https://api.bte.ncats.io/v1/query" --header "Content-Type: application/json" --data-raw "{
    \"message\": {
        \"query_graph\": {
            \"edges\": {
                \"e01\": {
                    \"object\": \"n0\",
                    \"subject\": \"n1\",
                    \"predicates\":[\"biolink:negatively_regulates_entity_to_entity\"]
                }
            },
            \"nodes\": {
                \"n0\": {
                    \"ids\":[\"NCBIGene:23221\"],
                    \"categories\": [
                        \"biolink:Gene\"
                    ]
                },
                \"n1\": {
                    \"categories\": [
                        \"biolink:Gene\"
                    ]
                }
            }
        }
    }
}"

cbizon commented 3 years ago

FWIW, the E2F1 control is probably not a good one. The support is coming from semmed, and the PMID it cites (Thanks @andrewsu for digging it up) is https://pubmed.ncbi.nlm.nih.gov/18039672/. Looking at it, it's actually saying I think that E2F1 is an up-regulator of RHOBTB2, not a downregulator.

I'd be more interested in the CMAP derived regulators from SPOKE.

dkoslicki commented 3 years ago

@andrewsu Ah, I see the issue now! We were sending TRAPI 1.0 formatted queries to that endpoint, unaware that it switched to TRAPI 1.1. We can fix that now. Thanks for the help!

andrewsu commented 3 years ago

@dkoslicki great! Our apologies for not communicating better and following a more orderly transition plan. We'll do better for 1.2!

dkoslicki commented 3 years ago

Just FYI: the Expander Agent issue with BTE has been resolved in RTX issue #1489

NCATSTranslator / testing

Gene "negatively regulates" RHOBTB2 (NCBIGene:23221) #68