Learning to Use the Force: Fitting Repulsive Potentials in Density-Functional Tight-Binding with Gaussian Process Regression.

08:00 EDT 10th March 2020 | BioPortfolio

Summary of "Learning to Use the Force: Fitting Repulsive Potentials in Density-Functional Tight-Binding with Gaussian Process Regression."

The Density-Functional Tight Binding (DFTB) method is a popular semiempirical approximation to Density Functional Theory (DFT). In many cases, DFTB can provide comparable accuracy to DFT at a fraction of the cost, enabling simulations on length- and time-scales that are unfeasible with first principles DFT. At the same time (and in contrast to empirical interatomic potentials and force-fields), DFTB still offers direct access to electronic properties such as the band-structure. These advantages come at the cost of introducing empirical parameters to the method, leading to a reduced transferability compared to true first-principle approaches. Consequently, it would be very useful if the parameter-sets could be routinely adjusted for a given project. While fairly robust and transferable parameterization workflows exist for the electronic structure part of DFTB, the so-called repulsive potential $V_{\mathrm{rep}}$ poses a major challenge. In this paper we propose a machine-learning (ML) approach to fitting $V_{\mathrm{rep}}$, using Gaussian Process Regression (GPR) \hl{to reconstruct} $V_{\mathrm{rep}}$ \hl{with DFT-DFTB force residues as training data.} The use of GPR circumvents the need for non-linear or global parameter optimization, while at the same time offering arbitrary flexibility in terms of the functional form. We also show that the proposed method can be applied to multiple elements at once, by fitting repulsive potentials for organic molecules containing carbon, hydrogen and oxygen. Overall, the new approach removes focus from the choice of functional form and parameterization procedure, in favour of a data-driven philosophy.


Journal Details

This article was published in the following journal.

Name: Journal of chemical theory and computation
ISSN: 1549-9626


DeepDyve research library

PubMed Articles [17162 Associated PubMed Articles listed on BioPortfolio]

Development of Density Functional Tight-Binding Parameters Using Relative Energy Fitting and Particle Swarm Optimization.

We provide a strategy to optimize density functional tight-binding (DFTB) parameterization for the calculation of the structures and properties of organic molecules consisting of hydrogen, carbon, nit...

Large-Scale Molecular Dynamics Simulation for Ground and Excited States based on Divide-and-Conquer Long-Range Corrected Density Functional Tight-Binding Method.

In this study, divide-and-conquer (DC)-based density-functional tight-binding (DFTB) and time-dependent density-functional tight-binding (TDDFTB) methods were developed using long-range correction (LC...

Flexible fitting of small-molecules into electron microscopy maps using molecular dynamics simulations with Neural Network Potentials.

Despite significant advances in resolution, the potential for cryo-electron microscopy (EM) to be used in determining the structures of protein-drug complexes remains unrealized. Determination of accu...

Accelerating time-dependent density functional theory and GW calculations for molecules and nanoclusters with symmetry adapted interpolative separable density fitting.

Computing integrals over orbital pairs is one of the most costly steps in many popular first-principles methods used by the quantum chemistry and condensed matter physics community. Here we employ a r...

Structures and Energetics of Neutral and Cationic pyrene Clusters.

The low energy structures of neutral and cationic pyrene clusters containing up to seven molecules are searched through a global exploration scheme combining Parallel Tempering Monte Carlo algorithm a...

Clinical Trials [6766 Associated Clinical Trials listed on BioPortfolio]

Pre-oxygenation With High-flow Nasal Cannula in Adults During Rapid Sequence Induction Anesthesia

It has been demonstrated that Transnasal Humidified Rapid Insufflation Ventilatory Exchange used during preoxygenation for emergency surgery is at least equally effective as preoxygenation...

A Comparison of a Currently Marketed Hearing Aid Programmed With Two Different Fitting Methods

This study is designed to investigate the benefits of the hearing aid in the laboratory and in daily life when fitted with two different methods. Subjective and objective evaluations will ...

Influence of Transcranial Direct Stimulation Current on Motor Learning of Young Adults

Motor learning occurs with structural and functional modifications in neural networks that meet a certain demand. The improvement of performance in diverse activities is a measure of learn...

Coupler-Based Hearing-Aid Fitting Approach for Experienced Users

The overall aim of the study is to compare the verification accuracy and hearing-aid outcomes between the traditional, fitting approach where the patient is present during the visit and a ...

PreOxygenation In RSI Anesthesia

Pre-oxygenation with high-flow nasal cannula oxygen has been evaluated in a limited number of studies and seems to be better than traditional preoxygenation with a tight fitting mask. Oxyg...

Medical and Biotech [MESH] Definitions

Areas of attractive or repulsive force surrounding MAGNETS.

Process in which individuals take the initiative, in diagnosing their learning needs, formulating learning goals, identifying resources for learning, choosing and implementing learning strategies and evaluating learning outcomes (Knowles, 1975)

Depolarization of membrane potentials at the SYNAPTIC MEMBRANES of target neurons during neurotransmission. Excitatory postsynaptic potentials can singly or in summation reach the trigger threshold for ACTION POTENTIALS.

The fitting and adjusting of artificial parts of the body. (From Stedman's, 26th ed)

The recorded electrical responses from nerve, muscle, SENSORY RECEPTOR, or area of the CENTRAL NERVOUS SYSTEM following stimulation. They range from less than a microvolt to several microvolts. The evoked potential can be auditory (EVOKED POTENTIALS, AUDITORY), somatosensory (EVOKED POTENTIALS, SOMATOSENSORY), visual (EVOKED POTENTIALS, VISUAL), or motor (EVOKED POTENTIALS, MOTOR), or other modalities that have been reported. Often used synonymously to event-related potentials which are associated with higher level cognitive processes.

Quick Search

DeepDyve research library

Searches Linking to this Article