Riemannian flow model predicts molecular crystal structures efficiently
Riemannian Flow Models with Reinforcement Learning for Molecular Crystal Structure Prediction
Machine Learning
Summary
Predicting how molecules arrange themselves in crystals is important because the structure affects material properties. The authors developed a new AI model called CG-OMatG that represents molecules as rigid bodies and learns their positions and orientations in crystals. They improve predictions by combining geometry-based generative modeling with reinforcement learning to favor low-energy, realistic structures. Their model works well on benchmark datasets and can help discover new materials faster.
What this means in practice
- •For materials engineers: Generate candidate molecular crystal structures with realistic packing for faster material design iterations.
- •For computational chemists: Use the model to screen polymorphs efficiently, focusing on low-energy, experimentally relevant crystal forms.
Authors
Thomas Egg, Harry Winston Sullivan, Maya M. Martirossyan, Philipp Höllmer, Cheng Zeng, Adrian Roitberg, Mingjie Liu, Richard Hennig, Sapna Sarupria, Ellad B. Tadmor, Stefano Martiniani
Abstract
Crystal structure governs material properties, making crystal structure prediction (CSP) a fundamental problem in materials science. Generative models are a promising approach for solving this problem, but the prevalence of polymorphism, coupled with large unit cells and complex packing geometry, makes the molecular CSP task challenging for existing models. To address this, we introduce Coarse-Grained Open Materials Generation (CG-OMatG), an equivariant Riemannian flow-based generative model. CG-OMatG predicts molecular crystal structures \textit{via} a coarse-grained, hierarchical representation. CG-OMatG treats molecules as rigid bodies---performing both inter- and intra-molecular message passing to construct a geometric representation for molecular packings---and learns to reconstruct molecule centroid positions, orientations, and lattice parameters, conditioned on chemical species and conformer geometry. We train the model on subsets of the Open Molecular Crystals (OMC25) and Cambridge Structural Database (CSD) datasets. Further, we fine-tune the model \textit{via} policy gradient reinforcement learning to steer the model towards generating low-energy candidate structures. We validate the generated structures on the CSP blind test benchmark, assessing agreement with experimentally determined crystals using COMPACK packing-similarity analysis. CG-OMatG exhibits strong performance for generative molecular crystal structure prediction, paving the way for accelerated polymorph screening and organic solid-state materials discovery.