A Robust Crystal Structure Prediction Method to Support Small Molecule Drug Development with Large Scale Validation and Prospective Studies

Crystal polymorphism is an important and fascinating aspect of solid state chemistry with far reaching implications in the pharmaceuticals, agrisciences, nutraceuticals, battery and aviation industries. Late appearing more stable polymorphs have caused numerous issues in the pharmaceutical industry. Experimental polymorph screening can be very expensive and time consuming, and sometimes may miss important low energy polymorphs due to its inability to exhaust all crystallization conditions. In this paper, we report a crystal structure prediction (CSP) method with unprecedented accuracy and efficiency, validated on a large and diverse dataset including 65 molecules with 135 experimentally found polymorphic forms. The method combines a novel systematic crystal packing search algorithm and the first use of machine learning force fields in a hierarchical crystal energy ranking. Our method not only reproduces all the experimentally known polymorphs, but also suggests new low energy polymorphs yet to be discovered by experiment that might pose potential risks to development of the currently known forms of these compounds. In addition, we also demonstrate, in two prospective drug development projects, how the method was used to accelerate clinical formulation design and derisk downstream processing. The high accuracy, reliability, and efficiency of our method with large scale validations and prospective studies position it for routine molecular crystal structure prediction in drug development.