Graph deep learning models, which incorporate a natural inductive bias for a collection of atoms, are of immense interest in materials science and chemistry. Here, we introduce the Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry. Built on top of the popular Deep Graph Library (DGL) and Python Materials Genomics (Pymatgen) packages, our intention is for MatGL to be an extensible ``batteries-included'' library for the development of advanced graph deep learning models for materials property predictions and interatomic potentials. At present, MatGL has efficient implementations for both invariant and equivariant graph deep learning models, including the Materials 3-body Graph Network (M3GNet), MatErials Graph Network (MEGNet), Crystal Hamiltonian Graph Network (CHGNet), TensorNet and SO3Net architectures. MatGL also includes a variety of pre-trained universal interatomic potentials (aka ``foundational materials models (FMM)'') and property prediction models are also included for out-of-box usage, benchmarking and fine-tuning. Finally, MatGL includes support for Pytorch Lightning for rapid training of models.
@article{arxiv.2503.03837,
title = {Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry},
author = {Tsz Wai Ko and Bowen Deng and Marcel Nassar and Luis Barroso-Luque and Runze Liu and Ji Qi and Elliott Liu and Gerbrand Ceder and Santiago Miret and Shyue Ping Ong},
journal= {arXiv preprint arXiv:2503.03837},
year = {2025}
}
Comments
50 pages, 13 figures including Manuscript and Supplementary Inoformation