JOINTLY is an algorithm for interpretable joint clustering of single-cell and single-nucleus RNA-seq, using kernel-based joint non-negative matrix factorization. JOINTLY effectively captures shared cell states, while robustly accounting for dataset-specific states. JOINTLY is available as an R package here and as a Python package here.
This repository holds scripts to reproduce the analyses presented in the manuscript.
Datasets, intermediary files and results are availiable for download here
For the white adipose tissue atlas (WATLAS), the model weights are availiable here and the atlas can be explored and downloaded here