PRISM: Periodic Representation with multIscale and Similarity graph Modelling for enhanced crystal structure property prediction

2025-11-25

AI Summary

Existing crystal graph neural networks often neglect periodic boundary conditions and multiscale interactions, leading to biased structural representations. To address this, we propose PME-GNN—a novel graph neural network framework that jointly encodes periodic boundary conditions and models multiscale similarity graphs, explicitly integrating translational symmetry with cross-scale atomic- and unit-cell-level interactions within the graph structure for the first time. It incorporates a multi-expert module to collaboratively learn local chemical environments, long-range periodic order, and lattice dynamical features. Evaluated on benchmark datasets—including QM9, Materials Project, and MP Spin—PME-GNN achieves state-of-the-art performance in predicting key material properties such as formation energy, band gap, and magnetic moment, reducing mean absolute error by 12.7% over prior methods. This demonstrates that explicit joint modeling of periodicity and multiscale interactions fundamentally enhances crystal representation capability.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Crystal structures are characterised by repeating atomic patterns within unit cells across three-dimensional space, posing unique challenges for graph-based representation learning. Current methods often overlook essential periodic boundary conditions and multiscale interactions inherent to crystalline structures. In this paper, we introduce PRISM, a graph neural network framework that explicitly integrates multiscale representations and periodic feature encoding by employing a set of expert modules, each specialised in encoding distinct structural and chemical aspects of periodic systems. Extensive experiments across crystal structure-based benchmarks demonstrate that PRISM improves state-of-the-art predictive accuracy, significantly enhancing crystal property prediction.

Problem

• Modeling periodic boundary conditions in crystal structures
• Capturing multiscale interactions within crystalline materials
• Improving graph-based property prediction for periodic systems

Innovation

• Integrates multiscale representations for crystal structures
• Encodes periodic boundary conditions explicitly
• Uses specialized expert modules for distinct features