Dialogue clarity

[Nusob888]

Description of feature

Dear Pertpy team,

Can I make a request for more clarification on the dialogue implementation? There is no mention of preprocessing in the vignette and it's clear now that the example appears to be log normalised data despite the adata having no signs of being log normalised (e.g. ad.uns['log1p']). Also, the nCounts_RNA appears to have been transformed?

I assume a new adata was made after log normalisation and the nCounts_RNA is a sum of the log normalised counts? or its possible it was scaled after somehow. Naturally this changes the output of dialogue hugely from using raw counts.