Objective

• Traditional approaches to in-silico drug discovery have been hampered by incomplete and inaccurate representation of molecules. • This leads to poorly performing predictive models of efficacy and safety and inability to fully probe all of the drug-like chemical space. • GTN’s unique and proprietary platform combines deeper featurization of molecules using quantum mechanical descriptors with cutting-edge generative machine learning models to quickly generate novel intellectual property at scale. • This presentation gives an overview of the GTN platform, highlights an industrial case study and identifies future opportunities for the approach.