Comparing the Accuracy of High-Dimensional Neural Network Potentials and the Systematic Molecular Fragmentation Method: A Benchmark Study for all-trans Alkanes

Abstract

Many approaches, which have been developed to express the potential energy of large systems, exploit the locality of the atomic interactions. A prominent example are fragmentation methods, in which quantum chemical calculations are carried out for overlapping small fragments of a given molecule that are then combined in a second step to yield the system's total energy. Here we compare the accuracy of the systematic molecular fragmentation approach with the performance of high-dimensional neural network (HDNN) potentials introduced by Behler and Parrinello. HDNN potentials are similar in spirit to the fragmentation approach in that the total energy is constructed as a sum of environment-dependent atomic energies, which are derived indirectly from electronic structure calculations. As a benchmark set we use all-trans alkanes containing up to eleven carbon atoms at the coupled cluster level of theory. These molecules have been chosen because they allow to extrapolate reliable reference energies for very long chains, enabling an assessment of the energies obtained by both methods for alkanes including up to 10 000 carbon atoms. We find that both methods predict high-quality energies with the HDNN potentials yielding smaller errors with respect to the coupled cluster reference.

References in corpus (3)

• Deep Learning in Neural Networks: An Overview
• How van der Waals interactions determine the unique properties of water
• Comparing the Accuracy of High-Dimensional Neural Network Potentials and the Systematic Molecular Fragmentation Method: A Benchmark Study for all-trans Alkanes

Cited by in corpus (17)

• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra
• Machine learning for electronically excited states of molecules
• WACSF - Weighted Atom-Centered Symmetry Functions as Descriptors in Machine Learning Potentials
• Combining SchNet and SHARC: The SchNarc machine learning approach for excited-state dynamics
• Inverse design of 3d molecular structures with conditional generative neural networks
• Predicting Electronic Structure Properties of Transition Metal Complexes with Neural Networks
• The Many-Body Expansion Combined with Neural Networks
• Tutorial: How to Train a Neural Network Potential
• Comparing the Accuracy of High-Dimensional Neural Network Potentials and the Systematic Molecular Fragmentation Method: A Benchmark Study for all-trans Alkanes
• Machine learning and excited-state molecular dynamics
• Atomic Permutationally Invariant Polynomials for Fitting Molecular Force Fields
• Predicting Properties of Periodic Systems from Cluster Data: A Case Study of Liquid Water
• Transferability of machine learning potentials: Protonated water neural network potential applied to the protonated water hexamer
• Constructing grids for molecular quantum dynamics using an autoencoder
• Size and Temperature Transferability of Direct and Local Deep Neural Networks for Atomic Forces
• A Hessian-Based Assessment of Atomic Forces for Training Machine Learning Interatomic Potentials