Mixture Models & Bayesian Differential Expression¶
This tutorial walks through using scribe to perform unsupervised cell-type
discovery with a two-component mixture model and Bayesian differential
expression analysis on a 50/50 mixture of Jurkat and HEK 293T cells from the
10x Genomics public dataset.
Pre-computed outputs
This notebook requires a GPU to run. All outputs shown below were
pre-computed and exported to static HTML. To re-run it yourself, clone
the repository and execute the notebook with marimo edit docs/tutorials/5050_jurkat_293t.py
on a GPU-enabled machine.