A Cartesian encoding graph neural network for crystal structure property prediction: application to thermal ellipsoid estimation

• DOI: 10.1039/d4dd00352g
• Publication Date: 2025-01-24T07:41:13Z
• Authors: Àlex Solé; Albert Mosella-Mo...; Joan Cardona; Silvia Gómez-Coca; Daniel Aravena; Eliseo Ruiz; Javier Ruiz-Hidalgo
• Keywords: FOS: Computer and information sciences; Computer Science - Machine Learning; Machine Learning (cs.LG)

CartNet efficiently predicts anisotropic displacement parameters (ADPs) for molecular crystal structures, reducing computational costs and outperforming tested methods while achieving state-of-the-art accuracy across diverse datasets.