search

NCATSTranslator / testing

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

ChemicalSubstance Celecoxib (PUBCHEM.COMPOUND:2662) - Gene COX1 (HGNC:9604) #79

Open MarkDWilliams opened 3 years ago

MarkDWilliams commented 3 years ago

Query: celecoxibNoHop PK: cf42e088-ef4a-45e1-a839-1562fb202901 PUBCHEM.COMPOUND:2662 HGNC:9604 Control: inhibits, negatively_regulates, decreases_activity_of etc Results Tracking Sheet

Updated 6/22/2021 image

edeutsch commented 3 years ago

I think the PK listed above is an older one; it does not match the screen shot.

MarkDWilliams commented 3 years ago

Ah, you are correct. My mistake. I have edited the original.

MarkDWilliams commented 3 years ago

A few tools return results for this, but my understanding is that a query with two identified nodes but an unspecified edge connecting them should return the edges that connect them which does not seem to be the case for any of the results returned.

vdancik commented 3 years ago

Indeed that's the case of MolePro - it ignores ids of the object node.

cbizon commented 3 years ago

It looks like Aragorn had a component fall down this morning. It's back up now and returning results for this query when called directly. If you rerun, you will hopefully get 4 answers back from aragorn corresponding to these predictes:

related_to affected_by molecularly_interacts_with decreases_activity_of

MarkDWilliams commented 3 years ago

I'm not sure I have the permissions to cancel the meeting for tomorrow, but given that Juneteenth was just recognized as a Federal holiday, the NCATS folks won't be able to make it. I plan to go ahead with this query for our standup on Tuesday the 22nd

bill-baumgartner commented 3 years ago

@sstemann & @MarkDWilliams, I am scheduled to be out of the office today so I won't be able to attend today's standup.

Initial thoughts on this query with regard to the Text Mining Provider: our chemicals are normalized to CHEBI currently, and proteins/genes to Protein Ontology identifiers. We are working on a solution to facilitate the protein/gene conflation issue as well as our use of species-non-specific protein identifiers (https://github.com/NCATSTranslator/Text-Mining-Provider-Roadmap/issues/81, https://github.com/NCATSTranslator/Text-Mining-Provider-Roadmap/issues/88). Our chemical normalization will likely remain CHEBI-specific going forward however, so this query will always return zero hits from our KGs as written. I will look into its ability to return results when using the CHEBI id for Celecoxib when I return to the office. One of our pipelines is overly aggressive when it comes to filtering potential false positives for chemicals, and I think that is in play here.

marcdubybroad commented 3 years ago

The Genetics KP doesn't have ChemicalSubstance data, so it returning 0 results is expected.

CaseyTa commented 3 years ago

We're not expecting any results from OpenPredict or COHD. I've removed assignment for myself and @arifx, but please let me know if we should remain assigned.

andrewsu commented 3 years ago

BTE returns a syntactically valid, non-zero response. But it is not executing the query as intended -- the fix is being tracked in https://github.com/biothings/BioThings_Explorer_TRAPI/issues/112

dkoslicki commented 3 years ago

Not sure why Expander agent is tagged in this one, due to returning results. Removing the tag, but LMK if there remains and issue for Expander agent

rajshruti18 commented 3 years ago

CAM-KP does not have data on 'Celecoxib' and COX1 gene; therefore, it returns 0 results.