Corpus size for taxa and genes not matching KB UI?

[wdahdul]

Question about the corpus numbers returned for taxa and genes. Shouldn't these match the KB UI numbers, which are 6184 taxa and 18537 genes?

Here's the output from rphenoscape:

> corpus_size("taxa")
[1] 785
> corpus_size("genes")
[1] 18662

[balhoff]

At least for taxa, these are counting different things. The similarity corpus is counting variation profiles resulting from our ancestral state procedure.

[hlapp]

Indeed they are counting different things. The corpus size for annotated-taxa is now reported though as

> corpus_size("annotated-taxa")
[1] 6533

whereas the KB website says 6540. The KB info query also returns 6540:

> get_KBinfo()
Annotated taxa: 6540
Annotated characters: 14555
Annotated matrices: 256
Annotated states: 28461
Build time: 2022-07-07 01:24:38 EDT

@balhoff any suggestions as to why the discrepancy?

And I'm not sure how to get the UI to display a corpus size for genes. Where did you find that, @wdahdul ?

[balhoff]

@hlapp at the moment the databases for the website and for the latest service API are different. This is because I have some pending work to do on the web UI to handle semantic similarity service changes.

[hlapp]

@balhoff any update here prior to the TraitFest?