Atomistic structure learning

Abstract

One endeavour of modern physical chemistry is to use bottom-up approaches to design materials and drugs with desired properties. Here we introduce an atomistic structure learning algorithm (ASLA) that utilizes a convolutional neural network to build 2D compounds and layered structures atom by atom. The algorithm takes no prior data or knowledge on atomic interactions but inquires a first-principles quantum mechanical program for physical properties. Using reinforcement learning, the algorithm accumulates knowledge of chemical compound space for a given number and type of atoms and stores this in the neural network, ultimately learning the blueprint for the optimal structural arrangement of the atoms for a given target property. ASLA is demonstrated to work on diverse problems, including grain boundaries in graphene sheets, organic compound formation and a surface oxide structure. This approach to structure prediction is a first step toward direct manipulation of atoms with artificially intelligent first principles computer codes.

References in corpus (11)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Automatic chemical design using a data-driven continuous representation of molecules
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• A real-space grid implementation of the Projector Augmented Wave method
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Machine Learning of Accurate Energy-Conserving Molecular Force Fields
• Deep Reinforcement Learning for De-Novo Drug Design
• Machine learning prediction errors better than DFT accuracy
• Understanding molecular representations in machine learning: The role of uniqueness and target similarity
• Data-driven learning of total and local energies in elemental boron

Cited by in corpus (15)

• Perspective on integrating machine learning into computational chemistry and materials science
• Generative Models for Automatic Chemical Design
• Atomistic Structure Learning Algorithm with surrogate energy model relaxation
• Machine Learning and Data-Driven Methods in Computational Surface and Interface Science
• Generative diffusion model for surface structure discovery
• Through the eyes of a descriptor: Constructing complete, invertible descriptions of atomic environments
• Accelerating the prediction of inorganic surfaces with machine learning interatomic potentials
• Reinforcement Learning for Molecular Design Guided by Quantum Mechanics
• Augmentation of Universal Potentials for Broad Applications
• Mean-Field Density Matrix Decompositions
• Surface Phase Diagrams from Nested Sampling
• Phase diagrams of polymer-containing liquid mixtures with a theory-embedded neural network
• Cascading symmetry constraint during machine learning-enabled structural search for sulfur induced Cu(111)-$(\sqrt{43}\times \sqrt{43})$ surface reconstruction
• Symmetry-Aware Actor-Critic for 3D Molecular Design
• Atomistic Global Optimization X: A Python package for optimization of atomistic structures