Beyond Structure: Invariant Crystal Property Prediction with Pseudo-Particle Ray Diffraction

2025-09-25

AI Summary

Traditional graph neural networks (GNNs) suffer from limited receptive fields and local encoding, hindering effective modeling of long-range atomic interactions in crystals—leading to representational ambiguity and biased property predictions. To address this, we propose PRDNet, a crystal representation framework that integrates pseudo-particle ray diffraction (PRD) features, jointly encoding real-space graph structure and reciprocal-space diffraction patterns while preserving full crystal symmetry. PRDNet explicitly captures global periodicity via learnable pseudo-particles that synthesize diffraction patterns, enhancing sensitivity to elemental identities and local chemical environments. Evaluated on three major benchmarks—Materials Project, JARVIS-DFT, and MatBench—PRDNet consistently outperforms existing GNN-based models, achieving state-of-the-art performance across diverse crystal property prediction tasks. This demonstrates the critical value of reciprocal-space information in advancing crystalline material representation learning.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Representation Learning for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks

Abstract

Crystal property prediction, governed by quantum mechanical principles, is computationally prohibitive to solve exactly for large many-body systems using traditional density functional theory. While machine learning models have emerged as efficient approximations for large-scale applications, their performance is strongly influenced by the choice of atomic representation. Although modern graph-based approaches have progressively incorporated more structural information, they often fail to capture long-term atomic interactions due to finite receptive fields and local encoding schemes. This limitation leads to distinct crystals being mapped to identical representations, hindering accurate property prediction. To address this, we introduce PRDNet that leverages unique reciprocal-space diffraction besides graph representations. To enhance sensitivity to elemental and environmental variations, we employ a data-driven pseudo-particle to generate a synthetic diffraction pattern. PRDNet ensures full invariance to crystallographic symmetries. Extensive experiments are conducted on Materials Project, JARVIS-DFT, and MatBench, demonstrating that the proposed model achieves state-of-the-art performance.

Problem

• Predicting crystal properties without exact quantum calculations
• Overcoming limitations of local atomic representations in ML models
• Capturing long-range atomic interactions and elemental variations

Innovation

• Uses reciprocal-space diffraction with graph representations
• Employs data-driven pseudo-particle for synthetic diffraction
• Ensures full invariance to crystallographic symmetries