In the published manuscript published in Nature Methods, it was demonstrated that MultiVI can be used on multi-omics data that include protein abundance measurements (e.g. CITE-seq). However, the tutorial on scvi-tools only shows its use on scRNA-seq and scATAC-seq data. My question is, if I want to use Multi-VI with protein dat, how should I proceed?
If you review the MultiVI documentation, you’ll see that there’s a protein key, which will point to your protein matrix, just like TotalVI. If you have questions, you can review the TotalVI tutorial and verify that the same logic and parameters work.