AI3SD Video: Accelerating structure prediction models for materials discovery

Abstract

The discovery of new functional materials can be guided by computational screening, particularly if the structure of a material can be reliably predicted from its chemical composition. For this application, we have been developing the use energy-structure-function maps [1], which summarise the crystal structures available to a given molecule and the relevant properties that are predicted for these structures. The use of these methods is still limited by the computational cost of crystal structure prediction (CSP). Most of the cost of CSP is associated with the calculation of the relative energies of predicted crystal structures using energy models that are sufficiently accurate to provide reliable energetic rankings. To speed up these methods, we have been developing machine learning approaches to predict high quality energies (e.g. from solid state density functional theory) from structures that have been generated with computationally efficient energy models [2-4]. The talk will discuss the performance of these methods, which use Gaussian Process Regression based on descriptors of local environments of atoms within crystal structures. I will also describe how these descriptors can be used to more quickly navigate the structure-property landscapes of molecular crystals [5] and how fast CSP can be applied to screen chemical space for the most promising molecules for a given application [6].
[1] Functional materials discovery using energy–structure–function maps, A. Pulido et al, Nature 2017, 543, 657.
[2] Machine learning for the structure–energy–property landscapes of molecular crystals, F. Musil, S. De, J. Yang, J. E. Campbell, G. M. Day and M Ceriotti, Chem. Sci. 2018, 9, 1289-1300.
[3] Machine-Learned Fragment-Based Energies for Crystal Structure Prediction, D. McDonagh, C.-K. Skylaris and G. M. Day, J. Chem. Theory Comput. 2019, 15, 2743–2758
[4] Multi-fidelity Statistical Machine Learning for Molecular Crystal Structure Prediction, O. Egorova, R. Hafizi, D. C. Woods and G. M. Day, J. Phys. Chem. A 2020, 124, 39, 8065–8078.
[5] Distributed Multi-Objective Bayesian Optimization for the Intelligent Navigation of Energy Structure Function Maps For Efficient Property Discovery, E. Pyzer-Knapp, G. M. Day, L. Chen, A. I. Cooper, ChemRxiv 2020, https://doi.org/10.26434/chemrxiv.13019960.v1
[6] Evolutionary chemical space exploration for functional materials: computational organic semiconductor discovery, C. Y. Cheng, J. E. Campbell and G. M. Day, Chem. Sci. 2020, 11, 4922-4933.

More information

Identifiers

ORCID for Graeme M. Day: orcid.org/0000-0001-8396-2771

ORCID for Samantha Kanza: orcid.org/0000-0002-4831-9489

ORCID for Jeremy G. Frey: orcid.org/0000-0003-0842-4302

ORCID for Mahesan Niranjan: orcid.org/0000-0001-7021-140X

Catalogue record

Export record

Altmetrics