Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics

Li, Chenghan; Voth, Gregory A.

doi:10.1021/acs.jctc.1c01085

Title: Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics

Abstract

Ab initio molecular dynamics (AIMD) has become one of the most popular and robust approaches for modeling complicated chemical, liquid, and material systems. However, the formidable computational cost often limits its widespread application in simulations of the largest-scale systems. Furthermore, the situation becomes even more severe in cases where the hydrogen nuclei may be better described as quantized particles using a path integral representation. Here, we present a computational approach that combines machine learning with recent advances in path integral contraction schemes, and we achieve a 2 orders of magnitude acceleration over direct path integral AIMD simulation while at the same time maintaining its accuracy.

Authors:

Li, Chenghan ^[1];

^[1]

Univ. of Chicago, IL (United States)

Publication Date:: Tue Jan 04 00:00:00 EST 2022

Research Org.:: Univ. of Chicago, IL (United States); University of Chicago, IL (United States)

Sponsoring Org.:: USDOE Office of Science (SC), Basic Energy Sciences (BES). Chemical Sciences, Geosciences & Biosciences Division; National Institute of General Medical Sciences (NIGMS); National Institutes of Health (NIH); USDOE Office of Science (SC), Basic Energy Sciences (BES)

OSTI Identifier:: 1865676

Alternate Identifier(s):: OSTI ID: 2337969

Grant/Contract Number:: SC0018648

Resource Type:: Accepted Manuscript

Journal Name:: Journal of Chemical Theory and Computation

Additional Journal Information:: Journal Volume: 18; Journal Issue: 2; Journal ID: ISSN 1549-9618

Publisher:: American Chemical Society

Country of Publication:: United States

Language:: English

Subject:: 37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY; Solvation; Electronic structure; Ab initio molecular dynamics; Chemical structure; Computer simulations

Citation Formats


                    Li, Chenghan, and Voth, Gregory A. Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics.  United States: N. p., 2022. 
Web.  doi:10.1021/acs.jctc.1c01085.

Copy to clipboard


                    Li, Chenghan, & Voth, Gregory A. Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics.  United States.  https://doi.org/10.1021/acs.jctc.1c01085

Copy to clipboard


                    Li, Chenghan, and Voth, Gregory A. Tue .  
"Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics".  United States.  https://doi.org/10.1021/acs.jctc.1c01085.  https://www.osti.gov/servlets/purl/1865676.

Copy to clipboard


                    
@article{osti_1865676,

  title        = {Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics},

  author       = {Li, Chenghan and Voth, Gregory A.},

  abstractNote = {Ab initio molecular dynamics (AIMD) has become one of the most popular and robust approaches for modeling complicated chemical, liquid, and material systems. However, the formidable computational cost often limits its widespread application in simulations of the largest-scale systems. Furthermore, the situation becomes even more severe in cases where the hydrogen nuclei may be better described as quantized particles using a path integral representation. Here, we present a computational approach that combines machine learning with recent advances in path integral contraction schemes, and we achieve a 2 orders of magnitude acceleration over direct path integral AIMD simulation while at the same time maintaining its accuracy.},

  doi          = {10.1021/acs.jctc.1c01085},

  journal      = {Journal of Chemical Theory and Computation},

  number       = 2,

  volume       = 18,

  place        = {United States},

  year         = {Tue Jan 04 00:00:00 EST 2022},

  month        = {Tue Jan 04 00:00:00 EST 2022}

}

Copy to clipboard

Journal Article:

Free Publicly Available Full Text

Accepted Manuscript (DOE)

Publisher's Version of Record

https://doi.org/10.1021/acs.jctc.1c01085

Other availability

Search WorldCat to find libraries that may hold this journal

Save / Share:

Export Metadata

Save to My Library

Works referenced in this record:

Ab initio molecular dynamics: basic concepts, current trends and novel applications
journal, December 2002

Tuckerman, Mark E.
Journal of Physics: Condensed Matter, Vol. 14, Issue 50
DOI: 10.1088/0953-8984/14/50/202

Efficient First-Principles Calculation of the Quantum Kinetic Energy and Momentum Distribution of Nuclei
journal, September 2012

Ceriotti, Michele; Manolopoulos, David E.
Physical Review Letters, Vol. 109, Issue 10
DOI: 10.1103/PhysRevLett.109.100604

Using Constrained Density Functional Theory to Track Proton Transfers and to Sample Their Associated Free Energy Surface
journal, September 2021

Li, Chenghan; Voth, Gregory A.
Journal of Chemical Theory and Computation, Vol. 17, Issue 9
DOI: 10.1021/acs.jctc.1c00609

Minimal Experimental Bias on the Hydrogen Bond Greatly Improves Ab Initio Molecular Dynamics Simulations of Water
journal, July 2020

Calio, Paul B.; Hocky, Glen M.; Voth, Gregory A.
Journal of Chemical Theory and Computation, Vol. 16, Issue 9
DOI: 10.1021/acs.jctc.0c00558

Accelerating Ab Initio Path Integral Simulations via Imaginary Multiple-Timestepping
journal, March 2016

Cheng, Xiaolu; Herr, Jonathan D.; Steele, Ryan P.
Journal of Chemical Theory and Computation, Vol. 12, Issue 4
DOI: 10.1021/acs.jctc.6b00021

The formulation of quantum statistical mechanics based on the Feynman path centroid density. I. Equilibrium properties
journal, April 1994

Cao, Jianshu; Voth, Gregory A.
The Journal of Chemical Physics, Vol. 100, Issue 7
DOI: 10.1063/1.467175

Nuclear quantum effects enter the mainstream
journal, February 2018

Markland, Thomas E.; Ceriotti, Michele
Nature Reviews Chemistry, Vol. 2, Issue 3
DOI: 10.1038/s41570-017-0109

Quantum Dynamics and Spectroscopy of Ab Initio Liquid Water: The Interplay of Nuclear and Electronic Quantum Effects
journal, March 2017

Marsalek, Ondrej; Markland, Thomas E.
The Journal of Physical Chemistry Letters, Vol. 8, Issue 7
DOI: 10.1021/acs.jpclett.7b00391

Accelerated path-integral simulations using ring-polymer interpolation
journal, December 2017

Buxton, Samuel J.; Habershon, Scott
The Journal of Chemical Physics, Vol. 147, Issue 22
DOI: 10.1063/1.5006465

The Role and Perspective of Ab Initio Molecular Dynamics in the Study of Biological Systems
journal, June 2002

Carloni, Paolo; Rothlisberger, Ursula; Parrinello, Michele
Accounts of Chemical Research, Vol. 35, Issue 6
DOI: 10.1021/ar010018u

Communication: Multiple-timestep ab initio molecular dynamics with electron correlation
journal, July 2013

Steele, Ryan P.
The Journal of Chemical Physics, Vol. 139, Issue 1
DOI: 10.1063/1.4812568

Decoding the spectroscopic features and time scales of aqueous proton defects
journal, June 2018

Napoli, Joseph A.; Marsalek, Ondrej; Markland, Thomas E.
The Journal of Chemical Physics, Vol. 148, Issue 22
DOI: 10.1063/1.5023704

Quantum–classical simulation methods for hydrogen transfer in enzymes: a case study of dihydrofolate reductase
journal, April 2004

Hammes-Schiffer, Sharon
Current Opinion in Structural Biology, Vol. 14, Issue 2
DOI: 10.1016/j.sbi.2004.03.008

Nuclear Quantum Effects in Water and Aqueous Systems: Experiment, Theory, and Current Challenges
journal, April 2016

Ceriotti, Michele; Fang, Wei; Kusalik, Peter G.
Chemical Reviews, Vol. 116, Issue 13
DOI: 10.1021/acs.chemrev.5b00674

Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties
journal, September 1998

Elstner, M.; Porezag, D.; Jungnickel, G.
Physical Review B, Vol. 58, Issue 11, p. 7260-7268
DOI: 10.1103/PhysRevB.58.7260

The Self-Consistent Charge Density Functional Tight Binding Method Applied to Liquid Water and the Hydrated Excess Proton: Benchmark Simulations
journal, May 2010

Maupin, C. Mark; Aradi, Bálint; Voth, Gregory A.
The Journal of Physical Chemistry B, Vol. 114, Issue 20
DOI: 10.1021/jp1010555

Accurate molecular dynamics and nuclear quantum effects at low cost by multiple steps in real and imaginary time: Using density functional theory to accelerate wavefunction methods
journal, February 2016

Kapil, V.; VandeVondele, J.; Ceriotti, M.
The Journal of Chemical Physics, Vol. 144, Issue 5
DOI: 10.1063/1.4941091

Benchmark Study of the SCC-DFTB Approach for a Biomolecular Proton Channel
journal, November 2013

Liang, Ruibin; Swanson, Jessica M. J.; Voth, Gregory A.
Journal of Chemical Theory and Computation, Vol. 10, Issue 1
DOI: 10.1021/ct400832r

DFTB3: Extension of the Self-Consistent-Charge Density-Functional Tight-Binding Method (SCC-DFTB)
journal, March 2011

Gaus, Michael; Cui, Qiang; Elstner, Marcus
Journal of Chemical Theory and Computation, Vol. 7, Issue 4
DOI: 10.1021/ct100684s

Ab initio molecular dynamics with nuclear quantum effects at classical cost: Ring polymer contraction for density functional theory
journal, February 2016

Marsalek, Ondrej; Markland, Thomas E.
The Journal of Chemical Physics, Vol. 144, Issue 5
DOI: 10.1063/1.4941093

How to remove the spurious resonances from ring polymer molecular dynamics
journal, June 2014

Rossi, Mariana; Ceriotti, Michele; Manolopoulos, David E.
The Journal of Chemical Physics, Vol. 140, Issue 23
DOI: 10.1063/1.4883861

Nuclear Quantum Effects Largely Influence Molecular Dissociation and Proton Transfer in Liquid Water under an Electric Field
journal, October 2020

Cassone, Giuseppe
The Journal of Physical Chemistry Letters, Vol. 11, Issue 21
DOI: 10.1021/acs.jpclett.0c02581

On the Quantum Nature of the Shared Proton in Hydrogen Bonds
journal, February 1997

Tuckerman, M. E.
Science, Vol. 275, Issue 5301
DOI: 10.1126/science.275.5301.817

An efficient ring polymer contraction scheme for imaginary time path integral simulations
journal, July 2008

Markland, Thomas E.; Manolopoulos, David E.
The Journal of Chemical Physics, Vol. 129, Issue 2
DOI: 10.1063/1.2953308

A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu
journal, April 2010

Grimme, Stefan; Antony, Jens; Ehrlich, Stephan
The Journal of Chemical Physics, Vol. 132, Issue 15
DOI: 10.1063/1.3382344

Ab initio path integral molecular dynamics: Basic ideas
journal, March 1996

Marx, Dominik; Parrinello, Michele
The Journal of Chemical Physics, Vol. 104, Issue 11
DOI: 10.1063/1.471221

Application of the SCC-DFTB Method to Neutral and Protonated Water Clusters and Bulk Water
journal, May 2011

Goyal, Puja; Elstner, Marcus; Cui, Qiang
The Journal of Physical Chemistry B, Vol. 115, Issue 20
DOI: 10.1021/jp202259c

Similar Records in DOE PAGES and OSTI.GOV collections:

Minimal Experimental Bias on the Hydrogen Bond Greatly Improves Ab Initio Molecular Dynamics Simulations of Water

Journal Article Calio, Paul B. ; Hocky, Glen M. ; Voth, Gregory A. - Journal of Chemical Theory and Computation

Experiment Directed Simulations (EDS) is a method within a class of techniques seeking to improve molecular simulations by minimally biasing the system Hamiltonian to reproduce certain experimental observables. In a previous application of EDS to ab initio molecular dynamics (AIMD) simulation based on electronic density functional theory (DFT), the AIMD simulations of water were biased to reproduce its experimentally derived solvation structure. In particular, by solely biasing the O-O pair correlation functions, other structural and dynamical properties that were not biased were improved. In this work, the hypothesis is tested that directly biasing the OH pair correlation (and hence themore »« less
https://doi.org/10.1021/acs.jctc.0c00558

Full Text Available
Accelerating Ab Initio Simulation via Nested Monte Carlo and Machine Learned Reference Potentials

Journal Article Jadrich, Ryan Bradley ; Leiding, Jeffery Allen - Journal of Physical Chemistry. B, Condensed Matter, Materials, Surfaces, Interfaces and Biophysical Chemistry

As a corollary of the rapid advances in computing, ab initio simulation is playing an increasingly important role in modeling materials at the atomic scale. Two strategies are possible, ab initio Monte Carlo (AIMC) and molecular dynamics (AIMD) simulation. The former benefits from exact sampling from the correct thermodynamic distribution, while the latter is typically more efficient with its collective all-atom coordinate updates. In this study, using a relatively simple test model comprised of helium and argon, we show that AIMC can be brought up to, and even above, the performance levels of AIMD via a hybrid nested sampling/machine learningmore »« less
https://doi.org/10.1021/acs.jpcb.0c03738

Full Text Available
Ab initio molecular dynamics with enhanced sampling in heterogeneous catalysis

Journal Article Piccini, GiovanniMaria ; Lee, Mal-Soon ; Yuk, Simuck F. ; ... - Catalysis Science and Technology

Ab initio molecular dynamics simulations combined with enhanced sampling techniques are becoming widespread methods to investigate chemical phenomena in catalytic systems. These techniques automatically include finite temperature effects, anharmonicity, and collective dynamics in their robust description of enthalpic and entropic contributions, which can have significant impact on reaction free energy landscapes. This contrasts with standard ab initio static approaches that are based on assessing reaction free energies from various coarse-grained descriptions of the reaction potential energy surface. Enhanced sampling ab initio molecular dynamics opens the way to first principles simulations of systems of increasing complexity like solid/liquid catalytic interfaces. Here,more »« less
https://doi.org/10.1039/d1cy01329g

Full Text Available
Machine Learning Classical Interatomic Potentials for Molecular Dynamics from First-Principles Training Data

Journal Article Chan, Henry ; Narayanan, Badri ; Cherukara, Matthew J. ; ... - Journal of Physical Chemistry. C

The ever-increasing power of modern super-computers, along with the availability of highly scalable atomistic simulation codes, has begun to revolutionize predictive modeling of materials. In particular, molecular dynamics (MD) has led to breakthrough advances in diverse fields, including tribology, catalysis, sensing, and nanoparticle self-assembly. Furthermore, recent integration of MD simulations with X-ray characterization has demonstrated promise in real-time 3-D characterization of materials on the atomic scale. The popularity of MD is driven by its applicability at disparate length/time scales, ranging from ab initio MD (hundreds of atoms and tens of picoseconds) to all-atom classical MD (millions of atoms over microseconds),more »« less
https://doi.org/10.1021/acs.jpcc.8b09917
Static and Dynamic Correlations in Water: Comparison of Classical Ab Initio Molecular Dynamics at Elevated Temperature with Path Integral Simulations at Ambient Temperature

Journal Article Li, Chenghan ; Paesani, Francesco ; Voth, Gregory A. - Journal of Chemical Theory and Computation

It is a common practice in ab initio molecular dynamics (AIMD) simulations of water to use an elevated temperature to overcome the overstructuring and slow diffusion predicted by most current density functional theory (DFT) models. The simulation results obtained in this distinct thermodynamic state are then compared with experimental data at ambient temperature based on the rationale that a higher temperature effectively recovers nuclear quantum effects (NQEs) that are missing in the classical AIMD simulations. In this work, we systematically examine the foundation of this assumption for several DFT models as well as for the many-body MB-pol model. Here, wemore »« less
https://doi.org/10.1021/acs.jctc.1c01223

Full Text Available

Similar Records

Title: Using Machine Learning to Greatly Accelerate Path Integral Ab Initio Molecular Dynamics

Abstract

Citation Formats

Ab initio molecular dynamics: basic concepts, current trends and novel applications journal, December 2002

Efficient First-Principles Calculation of the Quantum Kinetic Energy and Momentum Distribution of Nuclei journal, September 2012

Using Constrained Density Functional Theory to Track Proton Transfers and to Sample Their Associated Free Energy Surface journal, September 2021

Minimal Experimental Bias on the Hydrogen Bond Greatly Improves Ab Initio Molecular Dynamics Simulations of Water journal, July 2020

Accelerating Ab Initio Path Integral Simulations via Imaginary Multiple-Timestepping journal, March 2016

The formulation of quantum statistical mechanics based on the Feynman path centroid density. I. Equilibrium properties journal, April 1994

Nuclear quantum effects enter the mainstream journal, February 2018

Quantum Dynamics and Spectroscopy of Ab Initio Liquid Water: The Interplay of Nuclear and Electronic Quantum Effects journal, March 2017

Accelerated path-integral simulations using ring-polymer interpolation journal, December 2017

The Role and Perspective of Ab Initio Molecular Dynamics in the Study of Biological Systems journal, June 2002

Communication: Multiple-timestep ab initio molecular dynamics with electron correlation journal, July 2013

Decoding the spectroscopic features and time scales of aqueous proton defects journal, June 2018

Quantum–classical simulation methods for hydrogen transfer in enzymes: a case study of dihydrofolate reductase journal, April 2004

Nuclear Quantum Effects in Water and Aqueous Systems: Experiment, Theory, and Current Challenges journal, April 2016

Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties journal, September 1998

The Self-Consistent Charge Density Functional Tight Binding Method Applied to Liquid Water and the Hydrated Excess Proton: Benchmark Simulations journal, May 2010

Accurate molecular dynamics and nuclear quantum effects at low cost by multiple steps in real and imaginary time: Using density functional theory to accelerate wavefunction methods journal, February 2016

Benchmark Study of the SCC-DFTB Approach for a Biomolecular Proton Channel journal, November 2013

DFTB3: Extension of the Self-Consistent-Charge Density-Functional Tight-Binding Method (SCC-DFTB) journal, March 2011

Ab initio molecular dynamics with nuclear quantum effects at classical cost: Ring polymer contraction for density functional theory journal, February 2016

How to remove the spurious resonances from ring polymer molecular dynamics journal, June 2014

Nuclear Quantum Effects Largely Influence Molecular Dissociation and Proton Transfer in Liquid Water under an Electric Field journal, October 2020

On the Quantum Nature of the Shared Proton in Hydrogen Bonds journal, February 1997

An efficient ring polymer contraction scheme for imaginary time path integral simulations journal, July 2008

A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu journal, April 2010

Ab initio path integral molecular dynamics: Basic ideas journal, March 1996

Application of the SCC-DFTB Method to Neutral and Protonated Water Clusters and Bulk Water journal, May 2011

Ab initio molecular dynamics: basic concepts, current trends and novel applications
journal, December 2002

Efficient First-Principles Calculation of the Quantum Kinetic Energy and Momentum Distribution of Nuclei
journal, September 2012

Using Constrained Density Functional Theory to Track Proton Transfers and to Sample Their Associated Free Energy Surface
journal, September 2021

Minimal Experimental Bias on the Hydrogen Bond Greatly Improves Ab Initio Molecular Dynamics Simulations of Water
journal, July 2020

Accelerating Ab Initio Path Integral Simulations via Imaginary Multiple-Timestepping
journal, March 2016

The formulation of quantum statistical mechanics based on the Feynman path centroid density. I. Equilibrium properties
journal, April 1994

Nuclear quantum effects enter the mainstream
journal, February 2018

Quantum Dynamics and Spectroscopy of Ab Initio Liquid Water: The Interplay of Nuclear and Electronic Quantum Effects
journal, March 2017

Accelerated path-integral simulations using ring-polymer interpolation
journal, December 2017

The Role and Perspective of Ab Initio Molecular Dynamics in the Study of Biological Systems
journal, June 2002

Communication: Multiple-timestep ab initio molecular dynamics with electron correlation
journal, July 2013

Decoding the spectroscopic features and time scales of aqueous proton defects
journal, June 2018

Quantum–classical simulation methods for hydrogen transfer in enzymes: a case study of dihydrofolate reductase
journal, April 2004

Nuclear Quantum Effects in Water and Aqueous Systems: Experiment, Theory, and Current Challenges
journal, April 2016

Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties
journal, September 1998

The Self-Consistent Charge Density Functional Tight Binding Method Applied to Liquid Water and the Hydrated Excess Proton: Benchmark Simulations
journal, May 2010

Accurate molecular dynamics and nuclear quantum effects at low cost by multiple steps in real and imaginary time: Using density functional theory to accelerate wavefunction methods
journal, February 2016

Benchmark Study of the SCC-DFTB Approach for a Biomolecular Proton Channel
journal, November 2013

DFTB3: Extension of the Self-Consistent-Charge Density-Functional Tight-Binding Method (SCC-DFTB)
journal, March 2011

Ab initio molecular dynamics with nuclear quantum effects at classical cost: Ring polymer contraction for density functional theory
journal, February 2016

How to remove the spurious resonances from ring polymer molecular dynamics
journal, June 2014

Nuclear Quantum Effects Largely Influence Molecular Dissociation and Proton Transfer in Liquid Water under an Electric Field
journal, October 2020

On the Quantum Nature of the Shared Proton in Hydrogen Bonds
journal, February 1997

An efficient ring polymer contraction scheme for imaginary time path integral simulations
journal, July 2008

A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu
journal, April 2010

Ab initio path integral molecular dynamics: Basic ideas
journal, March 1996

Application of the SCC-DFTB Method to Neutral and Protonated Water Clusters and Bulk Water
journal, May 2011