Geometric Deep Learning for Molecular Crystal Structure Prediction

Michael Kilgour, Jutta Rogal, Mark Tuckerman

Research output: Contribution to journalArticlepeer-review


We develop and test new machine learning strategies for accelerating molecular crystal structure ranking and crystal property prediction using tools from geometric deep learning on molecular graphs. Leveraging developments in graph-based learning and the availability of large molecular crystal data sets, we train models for density prediction and stability ranking which are accurate, fast to evaluate, and applicable to molecules of widely varying size and composition. Our density prediction model, MolXtalNet-D, achieves state-of-the-art performance, with lower than 2% mean absolute error on a large and diverse test data set. Our crystal ranking tool, MolXtalNet-S, correctly discriminates experimental samples from synthetically generated fakes and is further validated through analysis of the submissions to the Cambridge Structural Database Blind Tests 5 and 6. Our new tools are computationally cheap and flexible enough to be deployed within an existing crystal structure prediction pipeline both to reduce the search space and score/filter crystal structure candidates.

Original languageEnglish (US)
Pages (from-to)4743-4756
Number of pages14
JournalJournal of chemical theory and computation
Issue number14
StatePublished - Jul 25 2023

ASJC Scopus subject areas

  • Computer Science Applications
  • Physical and Theoretical Chemistry


Dive into the research topics of 'Geometric Deep Learning for Molecular Crystal Structure Prediction'. Together they form a unique fingerprint.

Cite this