Potassium hydride is different from potassium ion

NCATSTranslator / Feedback

A repo for tracking gaps in Translator data and finding ways to fill them.

7 stars 0 forks source link

Potassium hydride is different from potassium ion #740

Open TranslatorIssueCreator opened 3 months ago

TranslatorIssueCreator commented 3 months ago

Type: Bug Report

URL: https://ui.test.transltr.io/main/results?l=Potassium%20Ion&i=PUBCHEM.COMPOUND:813&t=3&r=0&q=66f83ee1-3054-4d7a-8e7b-35def1c1a267

ARS PK: 66f83ee1-3054-4d7a-8e7b-35def1c1a267

Steps to reproduce:

query for what genes may be upregulated by potassium ion 4th result TNF shows results for potassium hydride

Screenshots:

sandrine-muller commented 3 months ago

I searched explicitely for potassium ion (K+) and get back results with potassium hydride (KH) which is a different chemical. Not sure it is a node norm issue?

sandrine-muller-research commented 3 months ago

Interestingly the PubChem card that shows up is for K+...

sandrine-muller-research commented 3 months ago

same issue for CSF2,ZGLP1, IL3, TAB2, TMC8, SCYL1, SMURF1, CD40, ATF7, ... see assets sheet for full list

gaurav commented 3 months ago

Re: Potassium specifically, it looks like we have some "good" cliques and some "bad" cliques:

PUBCHEM.COMPOUND:5462222 for the potassium atom (good clique, includes CHEBI:26216 and MESH:D011188)
PUBCHEM.COMPOUND:813 for the potassium ion (good clique, includes CHEBI:29103)
CHEBI:32589 "potassium hydride" is a BAD clique, mixing together potassium the element with potassium hydride.

I think that weird KH clique is why you're running into this issue. The good news is that CHEBI:32589 is much reduced in the next release of NodeNorm, so -- while it will take a while for this to make it into Prod -- it looks like a fix is on its way!

sandrine-muller commented 3 months ago

Thank you so much @gaurav !