Reinforcement Learning for Molecular Design Guided by Quantum Mechanics
This work addresses the challenge of de-novo molecular design for chemistry and drug discovery, representing an incremental improvement by extending existing graph-based methods to spatial coordinates.
The authors tackled the problem of automating molecular design by introducing a reinforcement learning formulation that operates in Cartesian coordinates, enabling the generation of a broader class of molecules and using quantum-mechanical properties for rewards, with their agent efficiently learning to solve tasks from scratch in an invariant state-action space.
Automating molecular design using deep reinforcement learning (RL) holds the promise of accelerating the discovery of new chemical compounds. Existing approaches work with molecular graphs and thus ignore the location of atoms in space, which restricts them to 1) generating single organic molecules and 2) heuristic reward functions. To address this, we present a novel RL formulation for molecular design in Cartesian coordinates, thereby extending the class of molecules that can be built. Our reward function is directly based on fundamental physical properties such as the energy, which we approximate via fast quantum-chemical methods. To enable progress towards de-novo molecular design, we introduce MolGym, an RL environment comprising several challenging molecular design tasks along with baselines. In our experiments, we show that our agent can efficiently learn to solve these tasks from scratch by working in a translation and rotation invariant state-action space.