Software

Pydex is an open-source Python package for model-based optimal experiment design originally developed by Dr. Kusumo during his PhD thesis as part of the PharmaSEL-Prosperity partnership.

Its focus is on designing maximally informative experimental campaigns for model parameter precision, using a discretization approach and optimizing over experimental efforts. It can account for model uncertainties to compute risk-averse experimental campaigns, and it can also be combined with feasible space sampling techniques such as the code DEUS to enable safe optimal experimental campaigns.

Sargent Centre researchers have developed one of a handful of world-class platforms for Crystal Structure Prediction for organic molecules. The software suite includes CrystalPredictor, CSO-RM and CSO-FM (formerly CrystalOptimizer) and CrystalEstimator. This software suite has been used successfully to model numerous pharmaceutical compounds and has been deployed by several research groups in blind tests of crystal structure prediction. It has been used to correctly predict several polymorphs of the highly-challenging compound ROY several years before their structures were uncovered experimentally.

For more information, email us on sargent.centre@imperial.ac.uk.