search

NCATSTranslator / Feedback

A repo for tracking gaps in Translator data and finding ways to fill them.
7 stars 0 forks source link

Include affinities in increase/decrease queries? #820

Open cbizon opened 2 months ago

cbizon commented 2 months ago

When we ask chemical increases/decreases gene type questions, we get a lot of lookup results, which is good. Many of these are coming from places like binding db. Many of these should include things like pIC50s. This kind of information would be incredibly valuable and should be 1) returned 2) displayed

cbizon commented 2 months ago

e.g. PK: a5731faf-7cd8-42f2-81ef-5cf6030931fa

For example: According to bindingdb (our primary source) ruxolitinib has a 3 nM IC50, which seems worth including even at the table level.

dnsmith124 commented 2 months ago

@Genomewide thoughts on the above?

gprice1129 commented 2 months ago

The UI team needs to discuss internally before assigned to a release.

Genomewide commented 2 months ago

I like the idea. And we should display that if returned in the trapi at least as part of the EPC.

Genomewide commented 2 months ago

@cbizon is that information returned in the attributes?

cbizon commented 2 months ago

Yes, although it's somewhat hit or miss depending on which source things are coming from: image

Genomewide commented 2 months ago

ok. we should look at adding this to EPC or other place. If it is edge related info, I think it should go in the EPC myself. If we put it there for some intermediate edges and not for the inferred or direct edge (putting at the table level) that could be confusing? But that is without giving it much thought yet.

Genomewide commented 2 months ago

@cbizon I tried to find this in the ARAX interface and the raw json and can not find this info. Was your screenshot from the ARAX, Unsecret, or ARS data? I can not find the pKi anywhere. I also searched the combined json from the ARS and don't see it or the 'affinity_parameter' either. Please help.