Functional and Learnable Cell dynamicS
Repository for the preprint: A scalable gene network model of regulatory dynamics in single cells.
We introduce FLeCS, a functional and learnable model of cell dynamics that incorporates gene network structure into coupled differential equations. FLeCS:
- accurately infers cell dynamics at scale
- provides improved functional insights into transcriptional mechanisms
- simulates single-cell trajectories
To quickly apply FLeCS to your own data, see the tutorial notebook at notebooks/Tutorial.ipynb.
The tutorial uses a publicly available myeloid differentiation dataset (download link provided in the notebook) and
should run in just a few minutes on a personal laptop.
You need to have Python 3.8 or newer installed on your system. If you don't have Python installed, we recommend installing Mambaforge.
To install the conda environment on GPU, please run
conda env create -f environment_gpu.yml
conda activate flecs
pip install -e .
To install the conda environment on CPU, please run
conda env create -f environment_cpu.yml
conda activate flecs
pip install -e .
This installation should only take a few minutes on a personal laptop.
All scripts used to generate figures are provided in folders FLeCS/figure<N>.
Please refer to the documentation.
If you use FLeCS in your research, please cite the following preprint:
@misc{bertin2025scalablegenenetworkmodel,
title={A scalable gene network model of regulatory dynamics in single cells},
author={Paul Bertin and Joseph D. Viviano and Alejandro Tejada-Lapuerta and Weixu Wang and Stefan Bauer and Fabian J. Theis and Yoshua Bengio},
year={2025},
eprint={2503.20027},
archivePrefix={arXiv},
primaryClass={q-bio.MN},
url={https://arxiv.org/abs/2503.20027},
}