Dr. Timothy Newhouse - Computationally Augmented Total Synthesis

Efficient syntheses of complex small molecules, such as bioactive natural products, often involve detailed retrosynthetic planning and experimental evaluation of speculative approaches. The central challenge of such plans is that experimental evaluation of high-risk strategies is resource intensive, as it entails iterative attempts at unsuccessful strategies. Along with the rapid development of cheminformatics and artificial intelligence, computer-aided synthetic planning  has emerged to address this challenge. Herein, we report a complementary strategy that combines innovative human-generated synthetic plans with robust computational prediction of the feasibility of key steps in the proposed synthesis.  A neural network model (NNET) was trained on a literature-based dataset (from Reaxys®) to predict the outcome of a generally disfavored transformation, the 6-endo-trig radical cyclization. The model performance was rigorously tested by experimental validation. Based on the virtual screening of potential substrates with our NNET model, optimal disconnection and structural modifications were chosen, resulting in 5-8 step syntheses of three clovane sesquiterpenoids. This work establishes how a machine learning model informs human design and guides multistep syntheses of complex small molecules.