(389c) Keynote Talk: Intelligent Cloud-Based Algorithms for Pharmaceutical Drug Development

Authors: 
Bellucci, M. A., XtalPi Inc
Burger, V., XtalPi Inc.
Yang, M., XtalPi Inc.
Sun, G., XtalPi Inc.
Liu, Y., XtalPi Inc.
Zhang, P., XtalPi Inc.
Ma, J., XtalPi Inc.
Jiang, A., XtalPi Inc.
Wen, S., XtalPi Inc.
Shi, X., XtalPi Inc
The ability to predict or explore polymorphism in organic crystal compounds is of great
importance for many industries since many physical and chemical properties of
molecular crystals strongly depend on the molecular packing in the solid state. Crystal
structure prediction (CSP) aims to reduce the risk and improve the efficiency of
pharmaceutical drug manufacturing by correctly predicting the low energy crystal forms
of organic molecules from minimal information, such as the chemical diagram. At XtalPi,
we have developed a cloud-based computational platform that combines advanced
physics-based algorithms with A.I/machine learning algorithms. Our CSP platform can
rapidly generate and rank millions of crystal structures for pharmaceutically relevant
systems and we highlight some cases in which our CSP platform has revealed low energy
polymorphs in challenging systems. We will also discuss some areas where we are
expanding our computational platform to make accurate predictions on the
physiochemical and pharmaceutical properties of small-molecule candidates for drug
design, solid-form selection, and for other critical aspects of drug development.