search

wikipathways / BioThings_Explorer_PFOCR_prioritization

PFOCR for prioritization/clustering of BioThings Explorer (BTE) TRAPI results
MIT License
0 stars 0 forks source link

Query examples to test #2

Open ayushi-agrawal-gladstone opened 1 year ago

ayushi-agrawal-gladstone commented 1 year ago

Try running the notebook for more queries for further testing. Examples:

  1. Original example results from #538:
  2. Liver injury query from Question of the Month
  3. Alzheimer's disease query from Question of the Month
ayushi-agrawal-gladstone commented 1 year ago

@khanspers Please share the queries above.

khanspers commented 1 year ago

Regarding liver injury, I think what is meant are these two: https://github.com/NCATSTranslator/testing/issues/176 https://github.com/NCATSTranslator/testing/issues/196

There are a lot of different queries in those two tickets. I tried the first of two posted by Colleen, which has only two BTE results, and one PFOCR pathway result. TRAPI results URL: https://arax.ncats.io/api/arax/v1.3/response/6b4550d4-c059-4fb8-857c-87319bd5b671

khanspers commented 1 year ago

Other example queries:

What treats Huntington's Disease? https://arax.ncats.io/api/arax/v1.3/response/a46d7144-718e-40dc-bde7-8733f38d9d04 BTE Results 684 TRAPI results CURIES 267 Figure results 13 CURIES represented by pathways 25

Query ``` { "message": { "query_graph": { "nodes": { "n0": { "ids": [ "MONDO:0007739" ], "categories": [ "biolink:Disease" ], "name": "Huntington disease" }, "n1": {} }, "edges": { "e0": { "subject": "n1", "object": "n0", "predicates":[ "biolink:treats"] } } } } } ```
AlexanderPico commented 1 year ago

@khanspers We're looking for one or two examples that particularly highlight the capability of this notebook (and pfocr pathways) to group results into biologically meaningful sets.

For example, we want a case where the first pathway groups n results together around one mechanism and the second pathway groups m results together around a second, distinct mechanism.

This would allow the researcher to quickly comprehend n+m results as really pointing to two interesting biological stories.

One could imagine this type of result for queries about Alzheimer's, etc. where there are multiple distinct mechanisms being researched.

khanspers commented 1 year ago

Here are some more queries to try (tested these in the notebook already). Im not sure either of them qualifies based on Alex's example above, but it would be interesting to see what others think.

  1. Drug-repurposing: Tamoxifen (CHEBI:41774) originally for cancer, re-purposed for Leishmaniasis (MONDO:0011989) Query: Tamoxifen (NamedEntity) - Gene - related_to - Leishmaniasis (Diseases) BTE results (from ARAX): 545 BTE response ID: af9b7f87-177d-4de4-902d-e3c273aabcc2

  2. Gene related to both Alzheimers (MONDO:0004975) AND Huntingtons (MONDO:0007739) Query: Gene - related_to - [Alzheimers, Huntingtons] BTE results (from ARAX): 6787 BTE response ID: 390baaea-0402-4524-a068-090b05f2de88