Symmetry-Aware Actor-Critic for 3D Molecular Design

Abstract

Automating molecular design using deep reinforcement learning (RL) has the potential to greatly accelerate the search for novel materials. Despite recent progress on leveraging graph representations to design molecules, such methods are fundamentally limited by the lack of three-dimensional (3D) information. In light of this, we propose a novel actor-critic architecture for 3D molecular design that can generate molecular structures unattainable with previous approaches. This is achieved by exploiting the symmetries of the design process through a rotationally covariant state-action representation based on a spherical harmonics series expansion. We demonstrate the benefits of our approach on several 3D molecular design tasks, where we find that building in such symmetries significantly improves generalization and the quality of generated molecules.

References in corpus (16)

• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• Objective-Reinforced Generative Adversarial Networks (ORGAN) for Sequence Generation Models
• MolGAN: An implicit generative model for small molecular graphs
• Learning Deep Generative Models of Graphs
• Constrained Graph Variational Autoencoders for Molecule Design
• SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks
• On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups
• 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data
• Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data
• Relevance of Rotationally Equivariant Convolutions for Predicting Molecular Properties
• MDP Homomorphic Networks: Group Symmetries in Reinforcement Learning
• Atomistic structure learning
• Comprehensive Analysis of the Neglect of Diatomic Differential Overlap Approximation
• Generating equilibrium molecules with deep neural networks

Cited by in corpus (4)

• SchNetPack 2.0: A neural network toolbox for atomistic machine learning
• An End-to-End Framework for Molecular Conformation Generation via Bilevel Programming
• Artificial Intelligence based Autonomous Molecular Design for Medical Therapeutic: A Perspective
• An In-depth Summary of Recent Artificial Intelligence Applications in Drug Design