search

NCATSTranslator / testing

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

Gene (NCBIGene:6656) - related_to - BiologicalProcessOrActivity - related_to - Gene (NCBIGene:6657) #111

Open sstemann opened 3 years ago

sstemann commented 3 years ago

Query: soxGenes.json PK: 597d7d54-e5b0-44be-8bf1-edf7d7fe73ce Control: DNA binding and transcription type stuff Results Tracking Sheet

image

cbizon commented 3 years ago

Looks like aragorn is returning results, including dna binding. There does seem to be a lot of duplication, tracking here: https://github.com/ranking-agent/strider/issues/267

marcdubybroad commented 3 years ago

The genetics KP doesn't handle multi hop queries, so the lack of results is expected.

bill-baumgartner commented 3 years ago

The text mining provider also does not handle multi-hop queries, so zero results are expected.

yakaboskic commented 3 years ago

CHP does not handle multi-hop queries and also doesn’t handle Biolink:BiologicalProcessOrActivity categories. However, we hope to support the latter soon as we begin to ingest pathway information in the next couple weeks. Zero results are expected.

CaseyTa commented 3 years ago

COHD and OpenPredict do not support multihop queries, or Gene or BiologicalProcessOrActivity categories.

colleenXu commented 2 years ago

BTE now handles the query correctly, after its long-awaited update! @andrewsu I think we can be unassigned from this issue.

(note that the link and screenshot are for TRAPI v1.1)

PK: 79a5e048-483c-4a69-8406-15ae6430a51a Screen Shot 2021-09-10 at 9 53 46 PM

More discussion of BTE's handling of this query here

GregHydeDartmouth commented 2 years ago

CHP will not handle this type of a query however one of the queries we handle follows a similar vein that I thought would be interesting to discuss here. Our gene->gene edge (using a fill operation) will build gene associations given the survival time of patients in our cohort. To build this relationships we actually use a set of sensitive drugs internal to us as a proxy. So you could technically consider this as a gene->drug->gene. In the query provided you use biolink:BiologicalProcessOrActivity as a proxy instead. I wanted to specify SOX1 (NCBIGene:6656) and use the fill operation to our other gene to see if we returned SOX2 as being sensitive. This was not the case, though we do return some results that I thought I would share.

The query I ran: { "message": { "query_graph": { "nodes": { "n0": { "ids": [ "NCBIGene:6656" ], "categories": [ "biolink:Gene" ] }, "n2": { "categories":[ "biolink:Gene" ] } }, "edges": { "e0": { "subject": "n2", "object": "n0", "predicates":["biolink:related_to"] } } } } }

And the results are here. These results are consolidated into ARAX and scored by them as well.