Assets not available

[PeteHaitch]

Tried analysing a human dataset stored as SingleCellExperiment .rds file and hit this error

Error: failed to download 'https://cors-proxy.aaron-lun.workers.dev/https%3A%2F%2Fgithub.qkg1.top%2Fkanaverse%2Fkana-special-features%2Freleases%2Fdownload%2Fv1.0.0%2Fhuman-mito-symbol.txt.gz' (404)

Tried repeating but disabled using default mitochondrial list and instead hit this error

Error: failed to download 'https://cors-proxy.aaron-lun.workers.dev/https%3A%2F%2Fgithub.qkg1.top%2FLTLA%2Fgesel-feedstock%2Freleases%2Fdownload%2Fgenes-v1.0.0%2Fhuman_symbol.tsv.gz' (404)

I can't share this dataset immediately but it uses Ensembl gene IDs (e.g., ENSG00000187999.4) if that helps.

[LTLA]

I see the problem - it's here.

Normally, this code will return the taxonomy ID. But in your case, the row names don't match to our expectations of Ensembl IDs because of the trailing .[0-9]. So it fails to make any guess, and bakana falls back to the incorrect "human" default.

For now, the simple solution is to strip off the .[0-9]. I should also fix the default to point to 9606, but that won't be the most appropriate solution for you, as the suffixed Ensembl IDs would not match to the strings in the various assets that kana uses for cross-referencing. I suppose we could also strip .[0-9]+ patterns but I am wary of fiddling too much with the input dataset.

[PeteHaitch]

Thanks, that's an easy fix for me to make.
Perhaps an error message if the row names are an unexpected format would help others.