Enhancing crystal structure prediction methods for flexible small molecule pharmaceuticals

Abstract

This thesis presents methods for crystal structure prediction, introducing techniques that improve the identification of conformations used as seeds in the search for crystal structures. The proposed approach enhances sampling of the conformational hypersurface, leading to superior initial structures that facilitate a more thorough exploration of conformational space. Additionally, a Monte Carlo simulated annealing method has been developed, integrating experimental and computational techniques to effectively determine crystal structures. This method has demonstrated success for rigid molecules and polymorphs under various conditions. A Monte Carlo refinement procedure has also been utilised to enable precise matching of crystal structure prediction datasets to experimental data for both flexible and rigid molecules. These methodologies hold promising potential for diverse applications in crystal structure prediction but require further research to ensure their robustness for practical workloads.

More information

Identifiers

ORCID for James Bramley: orcid.org/0000-0003-0215-0355

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

ORCID for Chris Skylaris: orcid.org/0000-0003-0258-3433

Catalogue record

Export record