Quantum-chemical insights from deep tensor neural networks

2016/09/27 by Kristof T. Schütt, Farhad Arbabzadah, Stefan Chmiela +3 · 1 voice · 1,506 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · Materials Science · Mathematics · Physics and Astronomy · #Artificial intelligence #Artificial neural network #Bioinformatics #Biology #Chemical space #Computational Drug Discovery Methods #Computer science #Deep learning #Drug discovery #Machine Learning in Materials Science #Mathematics #Molecule #Observable #Physics #Protein Structure and Dynamics #Quantum #Quantum chemical #Quantum mechanics #Statistical physics #Tensor (intrinsic definition) #physics.chem-ph

paper · pdf · doi:10.1038/ncomms13890

published in Nature Communications 8(1), 13890 (Nature Portfolio)

arxiv created 2016/11/07 · openalex publication_date 2017/01/09 · arxiv updated 2017/05/05 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

Abstract Learning from data has led to paradigm shifts in a multitude of disciplines, including web, text and image search, speech recognition, as well as bioinformatics. Can machine learning enable similar breakthroughs in understanding quantum many-body systems? Here we develop an efficient deep learning approach that enables spatially and chemically resolved insights into quantum-mechanical observables of molecular systems. We unify concepts from many-body Hamiltonians with purpose-designed deep tensor neural networks, which leads to size-extensive and uniformly accurate (1 kcal mol −1 ) predictions in compositional and configurational chemical space for molecules of intermediate size. As an example of chemical relevance, the model reveals a classification of aromatic rings with respect to their stability. Further applications of our model for predicting atomic energies and local chemical potentials in molecules, reliable isomer energies, and molecules with peculiar electronic structure demonstrate the potential of machine learning for revealing insights into complex quantum-chemical systems.

Citations

• ImageNet classification with deep convolutional neural networks

Cited by

• DScribe: Library of descriptors for machine learning in materials science
• On the use of chemical bonding descriptors in machine learning
• SchNet – A deep learning architecture for molecules and materials
• Operators in quantum machine learning: Response properties in chemical space
• Less is more: Sampling chemical space with active learning
• Coarse Graining Molecular Dynamics with Graph Neural Networks
• q-pac : A Python package for machine learned charge equilibration models
• A Unifying Review of Deep and Shallow Anomaly Detection
• Metapath-fused heterogeneous graph network for molecular property prediction
• Multi-body effects in a coarse-grained protein force field
• Analyzing Atomic Interactions in Molecules as Learned by Neural Networks
• A Transdisciplinary Review of Deep Learning Research and Its Relevance for Water Resources Scientists
• Neural Network-Based Interatomic Potential for the Study of Thermal and Mechanical Properties of Siliceous Zeolites
• Graph Neural Network Force Fields (GPTFF-mol) for Organic Molecules from Optimization Trajectories (OpenGEM26)
• QupKake: Integrating Machine Learning and Quantum Chemistry for Micro-p K a Predictions
• Machine Learning for Molecular Simulation
• Machine Learning of Reactive Potentials
• Neural Network Potentials: A Concise Overview of Methods
• A Comprehensive Survey on Graph Neural Networks
• Pushing the Accuracy Limit of Foundation Neural Network Models with Quantum Monte Carlo Forces and Path Integrals
• Learning Gradient Fields for Molecular Conformation Generation
• Covariant Compositional Networks For Learning Graphs
• Efficient prediction of 3D electron densities using machine learning
• Lagrangian Neural Networks
• Crash testing machine learning force fields for molecules, materials, and interfaces: model analysis in the TEA Challenge 2023
• SpreadGNN: Serverless Multi-task Federated Learning for Graph Neural Networks
• MoleculeNet: A Benchmark for Molecular Machine Learning
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules
• Rotation Invariant Graph Neural Networks using Spin Convolutions
• Quantum-centric machine learning for molecular dynamics
• Connecting the concepts of quantum state tomography and molecular representations for machine learning
• Atomistic Simulations of Oxide-Water Interfaces using Machine Learning Potentials
• ForceNet: A Graph Neural Network for Large-Scale Quantum Calculations
• Ab initio machine learning in chemical compound space
• Methods for Interpreting and Understanding Deep Neural Networks
• A machine-learning framework for accelerating spin-lattice relaxation simulations
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• Four Generations of High-Dimensional Neural Network Potentials
• QSAR without borders
• Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
• High-throughput quantum theory of atoms in molecules (QTAIM) for geometric deep learning of molecular and reaction properties
• Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces
• Accelerating Eigenvalue Dataset Generation via Chebyshev Subspace Filter
• Coder as Editor: Code-driven Interpretable Molecular Optimization
• A Survey on Explainable Artificial Intelligence (XAI): Towards Medical XAI
• Graph Neural Networks: A Review of Methods and Applications
• Message Passing on the Edge: Towards Scalable and Expressive GNNs
• SimPoly: Simulation of Polymers with Machine Learning Force Fields Derived from First Principles
• Ab-Initio Solution of the Many-Electron Schrödinger Equation with Deep Neural Networks
• Multi-View Graph Learning with Graph-Tuple
• Analysis of Atomistic Representations Using Weighted Skip-Connections
• Adaptive Canonicalization with Application to Invariant Anisotropic Geometric Networks
• Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation
• Enhancing Molecular Dipole Moment Prediction with Multitask Machine Learning
• QuFoundry: Generating Data with Quantum Properties for Quantum Machine Learning Utility
• Efficient Grand Canonical Global Optimization with On-the-fly-trained Machine-learning Interatomic Potentials
• Symmetry-adapted graph neural networks for constructing molecular dynamics force fields
• Physics-aware, probabilistic model order reduction with guaranteed stability
• PotentialNet for Molecular Property Prediction
• Beyond Numerical Hessians: Higher-Order Derivatives for Machine Learning Interatomic Potentials via Automatic Differentiation
• FeNNol: an Efficient and Flexible Library for Building Force-field-enhanced Neural Network Potentials
• OMNI-P2x universal neural network potential for excited-state simulations
• Message Passing Networks for Molecules with Tetrahedral Chirality
• Deep learning and the Schrödinger equation
• Model Compression and Hardware Acceleration for Neural Networks: A Comprehensive Survey
• E(n) Equivariant Graph Neural Networks
• Unsupervised Atomic Data Mining via Multi-Kernel Graph Autoencoders for Machine Learning Force Fields
• Understanding and Comparing Deep Neural Networks for Age and Gender Classification
• A Graph to Graphs Framework for Retrosynthesis Prediction
• Differentiable Molecular Simulations for Control and Learning
• GaAN: Gated Attention Networks for Learning on Large and Spatiotemporal Graphs
• Node-Equivariant Message Passing for Efficient and Accurate Machine Learning Interatomic Potentials
• Deep Learning on Attributed Graphs: A Journey from Graphs to Their Embeddings and Back
• Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• Fast Haar Transforms for Graph Neural Networks
• Multi-View Graph Neural Networks for Molecular Property Prediction
• Machine learning in next-generation AEM fuel cells: a systematic review
• Attention-Based Learning on Molecular Ensembles
• Self-Supervised Graph Transformer on Large-Scale Molecular Data
• Combining Graph Neural Networks and Mixed Integer Linear Programming for Molecular Inference under the Two-Layered Model
• Symmetry-driven graph neural networks
• Atomistic Graph Neural Networks for metals: Application to bcc iron
• Explainable Artificial Intelligence: Understanding, Visualizing and Interpreting Deep Learning Models
• Deeply learning molecular structure-property relationships using attention- and gate-augmented graph convolutional network
• Uncertainty quantification of molecular property prediction using Bayesian neural network models
• A universal graph deep learning interatomic potential for the periodic table
• Learning from Similarity Proportion Loss for Classifying Skeletal Muscle Recovery Stages
• Particle Swarm Based Hyper-Parameter Optimization for Machine Learned\n Interatomic Potentials
• Scalar Coupling Constant Prediction Using Graph Embedding Local Attention Encoder
• Machine-Learned Potentials for Solvation Modeling
• Deep Neural Network Computes Electron Densities and Energies of a Large Set of Organic Molecules Faster than Density Functional Theory (DFT)
• Gated Graph Recursive Neural Networks for Molecular Property Prediction
• A Deep Learning Potential for Accurate Shock Response Simulations in Tin
• Understanding the Capabilities of Molecular Graph Neural Networks in Materials Science Through Multimodal Learning and Physical Context Encoding
• Probing the Temporal Response of Liquid Water to a THz Pump Pulse Using Machine Learning-Accelerated Non-Equilibrium Molecular Dynamics
• Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules and Materials Based on Gaussian Moments
• Molecule Property Prediction and Classification with Graph Hypernetworks
• Learning Diverse Fashion Collocation by Neural Graph Filtering
• DeepSymmetry : Using 3D convolutional networks for identification of tandem repeats and internal symmetries in protein structures
• Approximation spaces of deep neural networks
• Deep learning in bioinformatics: introduction, application, and perspective in big data era
• Artificial Intelligence based Autonomous Molecular Design for Medical Therapeutic: A Perspective
• The evolution of machine learning potentials for molecules, reactions and materials
• Heterogeneous Molecular Graph Neural Networks for Predicting Molecule Properties
• Learning Universal Graph Neural Network Embeddings With Aid Of Transfer Learning
• Geometric learning of the conformational dynamics of molecules using dynamic graph neural networks
• OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Features
• Graph Kernels: State-of-the-Art and Future Challenges
• Demand Forecasting from Spatiotemporal Data with Graph Networks and Temporal-Guided Embedding
• Deep machine learning potentials for multicomponent metallic melts: development, predictability and compositional transferability
• Automated discovery of fundamental variables hidden in experimental data
• Molecular force fields with gradient-domain machine learning (GDML): Comparison and synergies with classical force fields
• Hierarchical modeling of molecular energies using a deep neural network
• Time-lagged autoencoders: Deep learning of slow collective variables for molecular kinetics
• Forecasting the outcome of spintronic experiments with Neural Ordinary Differential Equations
• Construction of Machine Learned Force Fields with Quantum Chemical Accuracy: Applications and Chemical Insights
• Structure-based sampling and self-correcting machine learning for accurate calculations of potential energy surfaces and vibrational levels
• Symmetrical Graph Neural Network for Quantum Chemistry with Dual Real and Momenta Space
• Adaptive coupling of a deep neural network potential to a classical force field
• Complex reaction processes in combustion unraveled by neural network-based molecular dynamics simulation
• An Overview of Computational Approaches for Interpretation Analysis
• Machine learning on neutron and x-ray scattering and spectroscopies
• "What is relevant in a text document?": An interpretable machine learning approach
• The domestic novels of E.M. Forster : The longest journey, Howards end, A passage to India
• Zero-shot Autonomous Microscopy for Scalable and Intelligent Characterization of 2D Materials

Discussions

• Quantum-Chemical Insights from Deep Tensor Neural Networks [hn, 55 points, 3 comments]

Related

• Quantum-chemical insights from interpretable atomistic neural networks
• Unifying machine learning and quantum chemistry -- a deep neural network for molecular wavefunctions
• Exploring Chemical Compound Space with Quantum-Based Machine Learning
• Ab initio machine learning in chemical compound space
• Physical machine learning outperforms "human learning" in Quantum Chemistry