SPONGE Simulation Package tOwards Next GEneration molecular modeling

Developed by the Yi Qin Gao group at Peking University, SPONGE continues to advance high-performance molecular dynamics, enhanced sampling, free energy calculation, and AI-enhanced molecular simulation.

How to Cite

If SPONGE and related tools support your research, please cite the corresponding software or method papers.

SPONGE

SPONGE is the core high-performance molecular simulation program, supporting molecular dynamics, enhanced sampling, free energy calculation, and acceleration across multiple hardware platforms.

SPONGE: A GPU-Accelerated Molecular Dynamics Package with Enhanced Sampling and AI-Driven Algorithms DOI: 10.1002/cjoc.202100456

MindSPONGE

MindSPONGE is an AI-enhanced SPONGE branch developed with the MindSpore neural network framework.

Artificial Intelligence Enhanced Molecular Simulations DOI: 10.1021/acs.jctc.3c00214

Xponge

Xponge is a Python package for pre- and post-processing molecular simulations.

Xponge: A Python package to perform pre- and post-processing of molecular simulations DOI: 10.21105/joss.04467

Development Timeline

2026

SPONGE 2.0 enters beta testing for next-generation molecular simulation workflows

2024

SPONGE 1.4 is released with expanded simulation features and documentation

2023

SPONGE 1.3 is released, advancing enhanced sampling, free energy calculation, and examples

2021

The SPONGE software paper introduces GPU-accelerated MD, enhanced sampling, and AI-driven algorithms