Dissecting tumor cell programs through group biology estimation in clinical single-cell transcriptomics
Nature Communications Paper | Analysis | AACR '23 Oral Presentation| Cite Us
BEANIE is a non-parameteric method for identification of differentially expressed gene signatures in clinical single-cell RNA-seq datasets. It can compare between two clinical groups of patients that share a subpopulation of cells, and calculates a biologically contextualized p-value (empirical p-value) and robustness for each gene signature. Tutorials can be found on the project wiki.
It is recommended to run BEANIE in a separate conda environment. To setup a new environment -
# Create conda environment
conda create --name beanie_env python=3.7
# Activate conda environment
conda activate beanie_env
BEANIE v1.0.0 can be installed directly via github -
pip install git+https://github.com/vanallenlab/beanie.git
Requirements: python v3.7 and above.
If you've found this work useful, please cite the following :
Johri, S., Bi, K., Titchen, B.M. et al. Dissecting tumor cell programs through group biology estimation in clinical single-cell transcriptomics. Nat Commun 16, 2090 (2025). https://doi.org/10.1038/s41467-025-57377-6
Please report issues directly to sjohri@g.harvard.edu.