Atomistic Line Graph Neural Network for Improved Materials Property Predictions

Abstract

Graph neural networks (GNN) have been shown to provide substantial performance improvements for atomistic material representation and modeling compared with descriptor-based machine learning models. While most existing GNN models for atomistic predictions are based on atomic distance information, they do not explicitly incorporate bond angles, which are critical for distinguishing many atomic structures. Furthermore, many material properties are known to be sensitive to slight changes in bond angles. We present an Atomistic Line Graph Neural Network (ALIGNN), a GNN architecture that performs message passing on both the interatomic bond graph and its line graph corresponding to bond angles. We demonstrate that angle information can be explicitly and efficiently included, leading to improved performance on multiple atomistic prediction tasks. We ALIGNN models for predicting 52 solid-state and molecular properties available in the JARVIS-DFT, Materials project, and QM9 databases. ALIGNN can outperform some previously reported GNN models on atomistic prediction tasks by up to 85% in accuracy with better or comparable model training speed.

References in corpus (5)

• BoltzTraP. A code for calculating band-structure dependent quantities
• A General-Purpose Machine Learning Framework for Predicting Properties of Inorganic Materials
• Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Molecules
• Molecular Mechanics-Driven Graph Neural Network with Multiplex Graph for Molecular Structures
• High-throughput search for magnetic topological materials using spin-orbit spillage, machine-learning and experiments

Cited by in corpus (97)

• Recent Advances and Applications of Deep Learning Methods in Materials Science
• A Universal Graph Deep Learning Interatomic Potential for the Periodic Table
• Scalable Parallel Algorithm for Graph Neural Network Interatomic Potentials in Molecular Dynamics Simulations
• A practical guide to machine learning interatomic potentials -- Status and future
• A critical examination of robustness and generalizability of machine learning prediction of materials properties
• Infusing Theory into Deep Learning for Interpretable Reactivity Prediction
• Accelerating computational materials discovery with artificial intelligence and cloud high-performance computing: from large-scale screening to experimental validation
• Unified Graph Neural Network Force-field for the Periodic Table
• On the redundancy in large material datasets: efficient and robust learning with less data
• Performance Assessment of Universal Machine Learning Interatomic Potentials: Challenges and Directions for Materials' Surfaces
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• A generative artificial intelligence framework based on a molecular diffusion model for the design of metal-organic frameworks for carbon capture
• Graph Neural Network Predictions of Metal Organic Framework CO2 Adsorption Properties
• Transferable E(3) equivariant parameterization for Hamiltonian of molecules and solids
• Designing High-Tc Superconductors with BCS-inspired Screening, Density Functional Theory and Deep-learning
• AI-driven inverse design of materials: Past, present and future
• Artificial Intelligence in Material Engineering: A review on applications of AI in Material Engineering
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• Probing out-of-distribution generalization in machine learning for materials
• JARVIS-Leaderboard: A Large Scale Benchmark of Materials Design Methods
• Inverse Design of Next-generation Superconductors Using Data-driven Deep Generative Models
• Spin-Dependent Graph Neural Network Potential for Magnetic Materials
• Developments and applications of the OPTIMADE API for materials discovery, design, and data exchange
• GPTFF: A high-accuracy out-of-the-box universal AI force field for arbitrary inorganic materials
• Autonomous decision making for solid-state synthesis of inorganic materials
• ChemNLP: A Natural Language Processing based Library for Materials Chemistry Text Data
• Toward Accurate Interpretable Predictions of Materials Properties within Transformer Language Models
• Recent progress in the JARVIS infrastructure for next-generation data-driven materials design
• Multimodal Foundation Models for Material Property Prediction and Discovery
• DenseGNN: universal and scalable deeper graph neural networks for high-performance property prediction in crystals and molecules
• An equivariant graph neural network for the elasticity tensors of all seven crystal systems
• ET-AL: Entropy-Targeted Active Learning for Bias Mitigation in Materials Data
• A Quantum-Chemical Bonding Database for Solid-State Materials
• A Deep-learning Model for Fast Prediction of Vacancy Formation in Diverse Materials
• Artificial Intelligence in Materials Science and Engineering: Current Landscape, Key Challenges, and Future Trajectorie
• Formula graph self-attention network for representation-domain independent materials discovery
• RydIQule: A Graph-based Paradigm for Modelling Rydberg and Atomic Systems
• InterMat: Accelerating Band Offset Prediction in Semiconductor Interfaces with DFT and Deep Learning
• Efficient first principles based modeling via machine learning: from simple representations to high entropy materials
• Accelerating the prediction of inorganic surfaces with machine learning interatomic potentials
• Quantifying Disorder One Atom at a Time Using an Interpretable Graph Neural Network Paradigm
• Fast and Accurate Prediction of Material Properties with Three-Body Tight-Binding Model for the Periodic Table
• Multi-task graph neural networks for simultaneous prediction of global and atomic properties in ferromagnetic systems
• Hardness and fracture toughness models by symbolic regression
• Data-driven Design of High Pressure Hydride Superconductors using DFT and Deep Learning
• InvDesFlow: An AI-driven materials inverse design workflow to explore possible high-temperature superconductors
• Deep learning of spectra: Predicting the dielectric function of semiconductors
• GPT-assisted learning of structure-property relationships by graph neural networks: Application to rare-earth doped phosphors
• Optimal pre-train/fine-tune strategies for accurate material property predictions
• Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
• LLM4Mat-Bench: Benchmarking Large Language Models for Materials Property Prediction
• AtomVision: A machine vision library for atomistic images
• Geometry-enhanced graph neural network for learning the smoothness of glassy dynamics from static structure
• Impact of Data Bias on Machine Learning for Crystal Compound Synthesizability Predictions
• Universal machine learning interatomic potentials poised to supplant DFT in modeling general defects in metals and random alloys
• Accelerating superconductor discovery through tempered deep learning of the electron-phonon spectral function
• The relationship between activated H2 bond length and adsorption distance on MXenes identified with graph neural network and resonating valence bond theory
• General time-reversal equivariant neural network potential for magnetic materials
• The JARVIS Infrastructure is All You Need for Materials Design
• Accelerating Material Property Prediction using Generically Complete Isometry Invariants
• Local structure, thermodynamics, and melting curve of boron phosphide at high pressures by deep learning-driven ab initio simulations
• Discovery of structure-property relations for molecules via hypothesis-driven active learning over the chemical space
• Material Property Prediction using Graphs based on Generically Complete Isometry Invariants
• End-to-end AI framework for interpretable prediction of molecular and crystal properties
• Optical materials discovery and design with federated databases and machine learning
• SynCoTrain: A Dual Classifier PU-learning Framework for Synthesizability Prediction
• Machine Learning Potentials for Alloys: A Detailed Workflow to Predict Phase Diagrams and Benchmark Accuracy
• Enhancing crystal structure prediction by combining computational and experimental data via graph networks
• Benchmarking CHGNet Universal Machine Learning Interatomic Potential Against DFT and EXAFS: Case of Layered WS2 and MoS2
• Coarse-grained crystal graph neural networks for reticular materials design
• HTESP (High-throughput electronic structure package): a Package for the high-throughput $ab$ $initio$ calculations
• PINK: physical-informed machine learning for lattice thermal conductivity
• EOSnet: Embedded Overlap Structures for Graph Neural Networks in Predicting Material Properties
• Superconductor discovery in the emerging paradigm of Materials Informatics
• Efficient Structure-Informed Featurization and Property Prediction of Ordered, Dilute, and Random Atomic Structures
• Materials discovery acceleration by using condition generative methodology
• A Cartesian Encoding Graph Neural Network for Crystal Structures Property Prediction: Application to Thermal Ellipsoid Estimation
• Discovery of sustainable energy materials via the machine-learned material space
• Lean CNNs for mapping electron charge density fields to material properties
• Machine-learning Guided Search for Phonon-mediated Superconductivity in Boron and Carbon Compounds
• Neural Structure Fields with Application to Crystal Structure Autoencoders
• Leveraging transfer learning for accurate estimation of ionic migration barriers in solids
• Perovskite-R1: a domain-specialized large language model for intelligent discovery of precursor additives and experimental design
• Search for thermodynamically stable ambient-pressure superconducting hydrides in GNoME database
• HTSC-2025: A Benchmark Dataset of Ambient-Pressure High-Temperature Superconductors for AI-Driven Critical Temperature Prediction
• Self-supervised Representations and Node Embedding Graph Neural Networks for Accurate and Multi-scale Analysis of Materials
• CHILI: Chemically-Informed Large-scale Inorganic Nanomaterials Dataset for Advancing Graph Machine Learning
• Bridging Text and Crystal Structures: Literature-driven Contrastive Learning for Materials Science
• FastCHGNet: Training one Universal Interatomic Potential to 1.5 Hours with 32 GPUs
• Do Graph Neural Networks Work for High Entropy Alloys?
• Machine learning-driven elasticity prediction in advanced inorganic materials via convolutional neural networks
• Predicting Many Crystal Properties via an Adaptive Transformer-based Framework
• Unified theory of atom-centered representations and message-passing machine-learning schemes
• A New Workflow for Materials Discovery Bridging the Gap Between Experimental Databases and Graph Neural Networks
• High-throughput study of kagome compounds in the AV3Sb5 family
• KAN-Enhanced Contrastive Learning Accelerating Crystal Structure Identification from XRD Patterns
• Self-Consistent Coulomb Interactions from Constrained Dynamical Mean-Field Theory