Include generic drug name as a parenthetic after the trade name of the drug, in a result list in Translator

[TranslatorIssueCreator]

Type: Suggestion

URL: https://ui.test.transltr.io/results?l=Alzheimer%27s&t=0&q=fa8a1e75-dd48-45cf-920c-9614e2950377

ARS PK: c3ee1172-587e-4f48-9771-fa5d1a770b97

Steps to reproduce:

Providing this as a bug report just so I can include "steps to reproduce".

Go to this query result page: https://ui.test.transltr.io/results?l=Alzheimer%27s&t=0&q=fa8a1e75-dd48-45cf-920c-9614e2950377

Scroll down to "Qi-Fu-Yin". This result is based on a single SemMedDB result (from two papers). Not sure if it should have a 94% confidence as a treatment for Alzheimer's Disease.

Screenshots:

[saramsey]

When Translator returns a drug result whose name is a trade name, like "Exelon", it would be helpful to include the USAN/WHO generic name for the drug (in this case, rivastigmine) as a parenthetic after the trade name. So instead of this:

Exelon

this:

Exelon (rivastigmine)

The benefit of this change would be that the suffix of the USAN/WHO generic name is semantically significant. For example, "rivastigmine" ends in "-stigmine" which means it's an acetylcholinesterase inhibitor. As another example, if a USAN/WHO generic name ends in "mab" we know it's a monoclonal antibody. If it ends in "nib" we know it's an RTK inhibitor. If it ends in "-statin" we know it is an inhibitor of 3-hydroxy-3-methylglutaryl-coenzyme-A reductase. That sort of thing. Especially for new drugs with sorta generic names like "Exelon" (if you google that, the first result is a company that is totally unrelated to the drug). Just an idea!

[gprice1129]

@newgene is this something the attribute server provides?

[newgene]

yes, I believe we do have these generic names.

[gprice1129]

The UI team is planning to implement this for the Lobster release.

[sierra-moxon]

@gprice1129 - can this be closed? :)

[gprice1129]

@sierra-moxon we didn't get to do this for Lobster. We plan on implementing something in the same spirit for Octopus or Eel.

[sierra-moxon]

from TAQA:

[Sui] Two "same" things with different names are currently a "hop" in other places. We wouldn't want this "generic" to be considered a "hop," just an alias. [Gus] 👍

[Gwen] -but do generic and commercial versions have different curies? shouldn’t they already be merged in node norm?

[Gaurav] - yep, let's test.

[gaurav]

NodeNorm doesn't really have a way to distinguish between generic names and commercial/brand names: our goal is to conflate generic names and commercial/brand names into a single clique when drug_chemical_conflate is on, and to give that clique a generic name as per #461. So, ideally, we'll end up implementing this ticket the other way around: NodeNorm will give you the generic name, and then the Annotator can provide a trade name.

If we do want to include commercial/brand names in NodeNorm, we will probably need to get those from RxCUI and include them in the properties we store about each clique, but we can only implement that in Hammerhead at the earliest.

[newgene]

Annotator does include many possible names for a chemical/drug (via MyChem.info) already, and we can probably work out a priority list collectively to determine a "canonical name", similarly how we define a priority list for the "canonical" ID field for a chemical.

[gaurav]

As per my previous comment, this should have been assigned to Hammerhead, not Guppy.