the wikidata metabolite ID list seems to incomplete

[tabbassidaloii]

I used an example in the app for the Wikidata metabolite ID (Q422964), but it cannot be found in the queried data.
Maybe we need to check the number of outputs as a quality control for the queried data?

[DeniseSl22]

Actually, this ID not existing does make sense, see:

I'm not query this type of relationship ("instance of" "type of a chemical entity").
I will need to check if this is linked to any of the other classes (then I could use the * operator I believe to retrieve all child-terms), and check if this annotation is actually correct.

[tabbassidaloii]

I see, thank you for the clarification

[DeniseSl22]

There are 13201 results in 30177 ms for "type of chemical entity".

I will first add that all subclasses need to be retrieved as well for all queries (with p:P31/ps:P31/wdt:P279*, see here.

If that doesn't resolve this issue, I will add the "type of chemical entity" as a separate relationship we're interested in (hoping it will not lead to timeouts 🤞 )

[tabbassidaloii]

@DeniseSl22 is there an update on this?

[tabbassidaloii]

@DeniseSl22 is there an update on this?

[DeniseSl22]

@tabbassidaloii : I've updated the query to include chemical compounds that are typed as "type of chemical entity" (something with RDF structure of Wikidata I didn't foresee.). So the file should now be complete, see https://github.com/sec2pri/mapping_preprocessing/actions/runs/21477939096 and #274 . I'll merge this into main. after also updating the queries for genes and proteins.