DiagVI tutorials

[michelta00]

Formatting

• My tutorial has only one top-level (#) header

Reproducibility

• My tutorial works on Google Colab
• My tutorial sets scvi.settings.seed = 0 at the beginning of the notebook
• My tutorial has been run and includes outputs (e.g. plots, tables)

Other

• Counts and normalized data should co-exist in the datasets, see the API overview for an example
• For scRNA-seq data, normalization should be counts per median library size and then log1p transformed -- if not, a reason should be given

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[ori-kron-wis]

A few things:

1. I wouldn't commit with warnings.filterwarnings("ignore")
2. See other comments from the scvi-tools diagvi PR. e.g : compute_foscttm into a different location
3. Add convergence curves
4. Validate with the most recent scvi-tools/scib-metrics versions (from main branch)
5. Validate both CPU/GPU runs (even a few epochs to see it works logically)
6. Perhaps we want to compare to TotalVI and TotalANVI?. Both of them are multimodal for RNA and Protein data

[review-notebook-app]

View / edit / reply to this conversation on ReviewNB

ori-kron-wis commented on 2026-03-24T13:27:43Z
----------------------------------------------------------------

Perhaps we want to compare to TotalVI and TotalANVI?. both of them multimodal for RNA and Protein data