Cormorant: Covariant Molecular Neural Networks

Abstract

We propose Cormorant, a rotationally covariant neural network architecture for learning the behavior and properties of complex many-body physical systems. We apply these networks to molecular systems with two goals: learning atomic potential energy surfaces for use in Molecular Dynamics simulations, and learning ground state properties of molecules calculated by Density Functional Theory. Some of the key features of our network are that (a) each neuron explicitly corresponds to a subset of atoms; (b) the activation of each neuron is covariant to rotations, ensuring that overall the network is fully rotationally invariant. Furthermore, the non-linearity in our network is based upon tensor products and the Clebsch-Gordan decomposition, allowing the network to operate entirely in Fourier space. Cormorant significantly outperforms competing algorithms in learning molecular Potential Energy Surfaces from conformational geometries in the MD-17 dataset, and is competitive with other methods at learning geometric, energetic, electronic, and thermodynamic properties of molecules on the GDB-9 dataset.

References in corpus (16)

• Neural Message Passing for Quantum Chemistry
• Geometric deep learning: going beyond Euclidean data
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• On representing chemical environments
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Deep Convolutional Networks on Graph-Structured Data
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• 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
• Gauge Equivariant Convolutional Networks and the Icosahedral CNN
• Bispectrum Inversion with Application to Multireference Alignment
• N-body Networks: a Covariant Hierarchical Neural Network Architecture for Learning Atomic Potentials
• Clebsch-Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network

Cited by in corpus (33)

• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• Robust active flow control over a range of Reynolds numbers using an artificial neural network trained through deep reinforcement learning
• Learning from Protein Structure with Geometric Vector Perceptrons
• Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Science
• E(n) Equivariant Graph Neural Networks
• Informing Geometric Deep Learning with Electronic Interactions to Accelerate Quantum Chemistry
• Theory for Equivariant Quantum Neural Networks
• Fast Uncertainty Estimates in Deep Learning Interatomic Potentials
• Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids
• Hierarchical, rotation-equivariant neural networks to select structural models of protein complexes
• Equivariant Graph Neural Networks for 3D Macromolecular Structure
• Rotation Invariant Graph Neural Networks using Spin Convolutions
• Predicting tensorial molecular properties with equivariant machine-learning models
• Completeness of Atomic Structure Representations
• Lorentz group equivariant autoencoders
• A community-powered search of machine learning strategy space to find NMR property prediction models
• Optimal radial basis for density-based atomic representations
• Data efficiency and extrapolation trends in neural network interatomic potentials
• No Headache for PIPs: A PIP Potential for Aspirin Outperforms Other Machine-Learned Potentials
• Multiresolution Graph Transformers and Wavelet Positional Encoding for Learning Hierarchical Structures
• Frame Averaging for Invariant and Equivariant Network Design
• A smooth basis for atomistic machine learning
• $\Delta$-model correction of Foundation Model based on the models own understanding
• Efficient ensemble uncertainty estimation in Gaussian Processes Regression
• Scaling Graph Neural Networks to Large Proteins
• HEroBM: a deep equivariant graph neural network for universal backmapping from coarse-grained to all-atom representations
• Unifying O(3) Equivariant Neural Networks Design with Tensor-Network Formalism
• GeoT: A Geometry-aware Transformer for Reliable Molecular Property Prediction and Chemically Interpretable Representation Learning
• Detect the Interactions that Matter in Matter: Geometric Attention for Many-Body Systems
• Neural Polarization: Toward Electron Density for Molecules by Extending Equivariant Networks
• Knowledge Distillation of a Protein Language Model Yields a Foundational Implicit Solvent Model
• HOAX: A Hyperparameter Optimization Algorithm Explorer for Neural Networks