Deep Neural Network Computes Electron Densities and Energies of a Large Set of Organic Molecules Faster than Density Functional Theory (DFT)

Abstract

Density functional theory (DFT) is one of the main methods in Quantum Chemistry that offers an attractive trade off between the cost and accuracy of quantum chemical computations. The electron density plays a key role in DFT. In this work, we explore whether machine learning - more specifically, deep neural networks (DNNs) - can be trained to predict electron densities faster than DFT. First, we choose a practically efficient combination of a DFT functional and a basis set (PBE0/pcS-3) and use it to generate a database of DFT solutions for more than 133,000 organic molecules from a previously published database QM9. Next, we train a DNN to predict electron densities and energies of such molecules. The only input to the DNN is an approximate electron density computed with a cheap quantum chemical method in a small basis set (HF/cc-VDZ). We demonstrate that the DNN successfully learns differences in the electron densities arising both from electron correlation and small basis set artifacts in the HF computations. All qualitative features in density differences, including local minima on lone pairs, local maxima on nuclei, toroidal shapes around C-H and C-C bonds, complex shapes around aromatic and cyclopropane rings and CN group, etc. are captured by the DNN. Accuracy of energy predictions by the DNN is ~ 1 kcal/mol, on par with other models reported in the literature, while those models do not predict the electron density. Computations with the DNN, including HF computations, take much less time that DFT computations (by a factor of ~20-30 for most QM9 molecules in the current version, and it is clear how it could be further improved).

References in corpus (12)

• SchNet - a deep learning architecture for molecules and materials
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Long-Range Corrected Hybrid Density Functionals with Improved Dispersion Corrections
• Less is more: sampling chemical space with active learning
• By-passing the Kohn-Sham equations with machine learning
• Machine learning prediction errors better than DFT accuracy
• WACSF - Weighted Atom-Centered Symmetry Functions as Descriptors in Machine Learning Potentials
• Transferability in Machine Learning for Electronic Structure via the Molecular Orbital Basis
• Guest Editorial: Special Topic on Data-enabled Theoretical Chemistry
• Metadynamics for Training Neural Network Model Chemistries: a Competitive Assessment
• Kinetic energy densities based on the fourth order gradient expansion: performance in different classes of materials and improvement via machine learning

Cited by in corpus (13)

• Using Attribution to Decode Dataset Bias in Neural Network Models for Chemistry
• Review: Deep Learning in Electron Microscopy
• Solving inverse-PDE problems with physics-aware neural networks
• Equivariant graph neural networks for fast electron density estimation of molecules, liquids, and solids
• PANNA: Properties from Artificial Neural Network Architectures
• Determining electronic properties from L-edge X-ray absorption spectra of transition metal compounds with artificial neural networks
• Mean-Field Density Matrix Decompositions
• Scalable Training of Trustworthy and Energy-Efficient Predictive Graph Foundation Models for Atomistic Materials Modeling: A Case Study with HydraGNN
• Efficient prediction of 3D electron densities using machine learning
• Embedding quantum statistical excitations in a classical force field
• Lean CNNs for mapping electron charge density fields to material properties
• On the equivalence of molecular graph convolution and molecular wave function with poor basis set
• Physical machine learning outperforms "human learning" in Quantum Chemistry