Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals

By Chi Chen, Weike Ye, Yunxing Zuo, Chen Zheng, Shyue Ping Ong. 2018

1,359 citations recorded by the source corpus.

Graph networks are a new machine learning (ML) paradigm that supports both relational reasoning and combinatorial generalization. Here, we develop, for the first time, universal MatErials Graph Network (MEGNet) models for accurate property prediction in both molecules and crystals. We demonstrate that our MEGNet models significantly outperform prior ML models in 11 out of 13 properties of the QM9 molecule data set. Furthermore, a single-task unified MEGNet model can accurately predict the internal energy at 0 K and room temperature, enthalpy and Gibbs free energy, with temperature, pressure and entropy being global state inputs. Similarly, we show that MEGNet models trained on $\sim 60,000$ crystals in the Materials Project substantially outperform prior ML models in the prediction of the formation energies, band gaps and elastic moduli of crystals, achieving better than DFT accuracy over a much larger data set. Such MEGNet models are highly interpretable, and well-established periodic chemical trends can be extracted from the elemental embeddings. Finally, we demonstrate the transfer learning of elemental embeddings from a property model trained on a larger data set (formation energies) to accelerate the training of property models with smaller amounts of data (band gaps and elastic moduli)

Subfields: Computational Theory and Mathematics, Materials Chemistry, Graph Learning, Computational Chemistry & Drug Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Materials Chemistry

In Graph Learning

In Computational Chemistry & Drug Discovery

Citations in the corpus

Directional Message Passing for Molecular Graphs (2020, methodology): “…2019), provably powerful graph networks (PPGN, results provided by the original authors) (Maron et al., 2019), MEGNet-simple (without auxiliary information) (Chen et al., 2019a), Cormorant (Anderson et al., 2019), and symmetrized gradient-domain machine learning (sGDML) (Chmiela et al., 2018).”
A universal graph deep learning interatomic potential for the periodic table (2022, methodology): “Such many-body calculations and graph convolutions can be repeated N times to construct models of arbitrary complexity, similar to previous materials graph network architectures 25 .”
A foundation model for atomistic materials chemistry. (2023, positive): “A key advance towards this goal was made by the MEGNet ( 21 ) model, which provided property prediction for inorganic crystals, and was trained on minimum energy configurations in the Materials Project (MP) ( 19 ) that includes most elements of the periodic table (89) and electronic structure…”
Atomistic Line Graph Neural Network for improved materials property predictions (2021, background): “…in the development of GNN architectures for predicting material properties such as SchNet 10 , Crystal Graph Convolutional Neural Networks (CGCNN) 15 , MatErials Graph Network (MEGNet) 16 , improved Crystal Graph Convolutional Neural Networks (iCGCNN) 17 , OrbNet 18 and similar variants 19-31 .”
Machine Learning for High-entropy Alloys: Progress, Challenges and Opportunities (2022, methodology): “Different GCNN models have been proposed, such as SchNet [240], MEGNet [241], DimeNet [242], and GemNet [242], and their implementation can be found in the Pytorch geometric package [243].”
Protein Representation Learning by Geometric Structure Pretraining (2022, methodology): “…existing methods employed node/atom message passing (Gilmer et al., 2017; Schütt et al., 2017a;b), edge/bond message passing (Jørgensen et al., 2018; Chen et al., 2019) and directional information (Klicpera et al., 2020; Liu et al., 2021; Klicpera et al., 2021) to encode 2D or 3D molecular graphs.”
Crystal Diffusion Variational Autoencoder for Periodic Material Generation (2021, methodology): “• Empirically, our model signiﬁcantly outperforms past methods in tasks including reconstructing an input structure, generating valid, diverse, and realistic materials, and generating materials that optimize speciﬁc properties. et al. (2019).”
Recent advances and applications of deep learning methods in materials science (2021, background): “com/materialsvirtuallab/megnet [33] DimeNet Mol https://github.”