革新知能統合研究センター

Speaker: Dr. Xiufeng Yang

Title: Monte Carlo Tree Search for Molecular Discovery

Abstract:
Molecule design with desired properties has important applications for creating novel functional materials and drugs. Recent advances on machine learning especially deep learning (generative models) allow us using data-driven approaches to automatically design novel molecules. And automatic design of organic materials requires black-box optimization in a vast chemical space. In conventional molecule design algorithms, a molecule is built as a combination of predetermined fragments. Recently, deep neural network models such as variational auto encoders (VAEs) and recurrent neural networks (RNNs) are shown to be effective in de novo design of molecules without any predetermined fragments. In our research, we introduce a novel python library ChemTS that explores the chemical space by combining MCTS and an RNN. One molecule generator is trained by RNN and MCTS is employed to discover better molecules by taking RNN as policy network. In a benchmarking problem of optimizing the octanol-water partition coefficient and synthesizability, our algorithm showed superior efficiency in finding high-scoring molecules.

To assess the reliability of AI-assisted chemistry, we prepared a platform using a molecule generator(ChemTS) and a DFT simulator, and attempted to generate novel photo-functional molecules whose lowest excited states lie at desired energetic levels. A ten-day run on 12 cores discovered 86 potential photo-functional molecules around target lowest excitation levels, designated as 200, 300, 400, 500, and 600nm. Among the molecules discovered, six were synthesized and five were confirmed to reproduce DFT predictions in ultraviolet visible absorption measurements. This result shows the potential of AI-assisted chemistry to discover ready-to-synthesize novel molecules.