A Python library for simulating gene regulatory networks (GRNs) using parameter-agnostic frameworks like RACIPE and Ising formalism, with GPU acceleration and efficient ODE solving.
Modeling gene regulatory networks (GRNs) is essential for understanding cellular processes, but parameterizing these networks becomes increasingly difficult as they scale. This Python library provides a simulation framework that unifies parameter-agnostic approaches, including RACIPE and Ising formalism, into a single, flexible tool.
- Simulation Frameworks: Supports both ODE-based (RACIPE) and coarse-grained (Ising formalism) methods for studying GRN dynamics.
- Parameter-Agnostic Modeling: Translates network topology into mathematical models without requiring detailed parameter tuning.
- Scalable Computation: Uses the Jax ecosystem for GPU acceleration and Diffrax for efficient ODE solving.
- Data Processing Tools: Provides normalization and discretization functions to standardize simulation outputs for downstream analysis.
You can access the full documentation, including installation instructions, usage examples, and detailed explanations of the simulation frameworks, at MoltenEcdysone09.github.io/GRiNS
The code to reproduce the analysis done in the manuscript can been accessed in the Case Studies folder. CS1 folder has the simulation and analysis reproduction scripts along with the raw and processed data for the first case study and CS2 has the same but for the second case study.
For optimal performance, it is recommended to install the GPU-accelerated version of the library. This version leverages CUDA for faster computations, making it well-suited for large-scale simulations. If you have a compatible NVIDIA GPU (refer to Jax Installation), install the library with:
pip install grins[cuda12]If you do not have a compatible GPU, you can install the CPU version instead:
pip install grinsCompared to the GPU version, the CPU version will be slower, especially for large simulations.
Please cite this package if you have used it.
@misc{harlapur2025grinspythonlibrarysimulating,
title={GRiNS: A Python Library for Simulating Gene Regulatory Network Dynamics},
author={Pradyumna Harlapur and Harshavardhan B V and Mohit Kumar Jolly},
year={2025},
eprint={2503.18356},
archivePrefix={arXiv},
primaryClass={q-bio.QM},
url={https://arxiv.org/abs/2503.18356},
}