Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry

Abstract Graph deep learning models, which incorporate a natural inductive bias for atomic structures, 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. B...

Full description

Saved in:
Bibliographic Details
Main Authors: Tsz Wai Ko, Bowen Deng, Marcel Nassar, Luis Barroso-Luque, Runze Liu, Ji Qi, Atul C. Thakur, Adesh Rohan Mishra, Elliott Liu, Gerbrand Ceder, Santiago Miret, Shyue Ping Ong
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:npj Computational Materials
Online Access:https://doi.org/10.1038/s41524-025-01742-y
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849331834628866048
author Tsz Wai Ko
Bowen Deng
Marcel Nassar
Luis Barroso-Luque
Runze Liu
Ji Qi
Atul C. Thakur
Adesh Rohan Mishra
Elliott Liu
Gerbrand Ceder
Santiago Miret
Shyue Ping Ong
author_facet Tsz Wai Ko
Bowen Deng
Marcel Nassar
Luis Barroso-Luque
Runze Liu
Ji Qi
Atul C. Thakur
Adesh Rohan Mishra
Elliott Liu
Gerbrand Ceder
Santiago Miret
Shyue Ping Ong
author_sort Tsz Wai Ko
collection DOAJ
description Abstract Graph deep learning models, which incorporate a natural inductive bias for atomic structures, 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, MatGL is designed to be an extensible “batteries-included” library for developing advanced model architectures 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 provides several pre-trained foundation potentials (FPs) with coverage of the entire periodic table, and property prediction models for out-of-box usage, benchmarking and fine-tuning. Finally, MatGL integrates with PyTorch Lightning to enable efficient model training.
format Article
id doaj-art-bb22674daaff4ea691d10638ef28cb56
institution Kabale University
issn 2057-3960
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series npj Computational Materials
spelling doaj-art-bb22674daaff4ea691d10638ef28cb562025-08-20T03:46:23ZengNature Portfolionpj Computational Materials2057-39602025-08-0111111410.1038/s41524-025-01742-yMaterials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistryTsz Wai Ko0Bowen Deng1Marcel Nassar2Luis Barroso-Luque3Runze Liu4Ji Qi5Atul C. Thakur6Adesh Rohan Mishra7Elliott Liu8Gerbrand Ceder9Santiago Miret10Shyue Ping Ong11Aiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoDepartment of Materials Science and Engineering, University of California BerkeleyIntel LabsDepartment of Materials Science and Engineering, University of California BerkeleyAiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoAiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoAiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoAiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoAiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoDepartment of Materials Science and Engineering, University of California BerkeleyIntel LabsAiiso Yufeng Li Family Department of Chemical and Nano Engineering, University of California San DiegoAbstract Graph deep learning models, which incorporate a natural inductive bias for atomic structures, 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, MatGL is designed to be an extensible “batteries-included” library for developing advanced model architectures 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 provides several pre-trained foundation potentials (FPs) with coverage of the entire periodic table, and property prediction models for out-of-box usage, benchmarking and fine-tuning. Finally, MatGL integrates with PyTorch Lightning to enable efficient model training.https://doi.org/10.1038/s41524-025-01742-y
spellingShingle Tsz Wai Ko
Bowen Deng
Marcel Nassar
Luis Barroso-Luque
Runze Liu
Ji Qi
Atul C. Thakur
Adesh Rohan Mishra
Elliott Liu
Gerbrand Ceder
Santiago Miret
Shyue Ping Ong
Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry
npj Computational Materials
title Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry
title_full Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry
title_fullStr Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry
title_full_unstemmed Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry
title_short Materials Graph Library (MatGL), an open-source graph deep learning library for materials science and chemistry
title_sort materials graph library matgl an open source graph deep learning library for materials science and chemistry
url https://doi.org/10.1038/s41524-025-01742-y
work_keys_str_mv AT tszwaiko materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT bowendeng materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT marcelnassar materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT luisbarrosoluque materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT runzeliu materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT jiqi materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT atulcthakur materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT adeshrohanmishra materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT elliottliu materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT gerbrandceder materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT santiagomiret materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry
AT shyuepingong materialsgraphlibrarymatglanopensourcegraphdeeplearninglibraryformaterialsscienceandchemistry