Crystal Structure Prediction by Joint Equivariant Diffusion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jiao, Rui, Huang, Wenbing, Lin, Peijia, Han, Jiaqi, Chen, Pin, Lu, Yutong, Liu, Yang
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909130040541184
author Jiao, Rui
Huang, Wenbing
Lin, Peijia
Han, Jiaqi
Chen, Pin
Lu, Yutong
Liu, Yang
author_facet Jiao, Rui
Huang, Wenbing
Lin, Peijia
Han, Jiaqi
Chen, Pin
Lu, Yutong
Liu, Yang
contents Crystal Structure Prediction (CSP) is crucial in various scientific disciplines. While CSP can be addressed by employing currently-prevailing generative models (e.g. diffusion models), this task encounters unique challenges owing to the symmetric geometry of crystal structures -- the invariance of translation, rotation, and periodicity. To incorporate the above symmetries, this paper proposes DiffCSP, a novel diffusion model to learn the structure distribution from stable crystals. To be specific, DiffCSP jointly generates the lattice and atom coordinates for each crystal by employing a periodic-E(3)-equivariant denoising model, to better model the crystal geometry. Notably, different from related equivariant generative approaches, DiffCSP leverages fractional coordinates other than Cartesian coordinates to represent crystals, remarkably promoting the diffusion and the generation process of atom positions. Extensive experiments verify that our DiffCSP significantly outperforms existing CSP methods, with a much lower computation cost in contrast to DFT-based methods. Moreover, the superiority of DiffCSP is also observed when it is extended for ab initio crystal generation.
format Preprint
id arxiv_https___arxiv_org_abs_2309_04475
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Crystal Structure Prediction by Joint Equivariant Diffusion
Jiao, Rui
Huang, Wenbing
Lin, Peijia
Han, Jiaqi
Chen, Pin
Lu, Yutong
Liu, Yang
Materials Science
Machine Learning
Crystal Structure Prediction (CSP) is crucial in various scientific disciplines. While CSP can be addressed by employing currently-prevailing generative models (e.g. diffusion models), this task encounters unique challenges owing to the symmetric geometry of crystal structures -- the invariance of translation, rotation, and periodicity. To incorporate the above symmetries, this paper proposes DiffCSP, a novel diffusion model to learn the structure distribution from stable crystals. To be specific, DiffCSP jointly generates the lattice and atom coordinates for each crystal by employing a periodic-E(3)-equivariant denoising model, to better model the crystal geometry. Notably, different from related equivariant generative approaches, DiffCSP leverages fractional coordinates other than Cartesian coordinates to represent crystals, remarkably promoting the diffusion and the generation process of atom positions. Extensive experiments verify that our DiffCSP significantly outperforms existing CSP methods, with a much lower computation cost in contrast to DFT-based methods. Moreover, the superiority of DiffCSP is also observed when it is extended for ab initio crystal generation.
title Crystal Structure Prediction by Joint Equivariant Diffusion
topic Materials Science
Machine Learning
url https://arxiv.org/abs/2309.04475