Machine Learning Force Field Parameters from Ab Initio Data

Li, Ying; Li, Hui; Pickard, Frank C.; Narayanan, Badri; Sen, Fatih G.; Chan, Maria Y.; Sankaranarayanan, Subramanian S.; Brooks, Bernard R.; Roux, Benoît

doi:10.1021/acs.jctc.7b00521

Title: Machine Learning Force Field Parameters from Ab Initio Data

Journal Article · 10 August 2017 · Journal of Chemical Theory and Computation

DOI: https://doi.org/10.1021/acs.jctc.7b00521 · OSTI ID:1415482

Li, Ying ^[1]; Li, Hui ^[2];

^[3]; Narayanan, Badri ^[4]; Sen, Fatih G. ^[4];

^[5]; Sankaranarayanan, Subramanian S. ^[5]; Brooks, Bernard R. ^[3];

^[6]

Argonne National Laboratory (ANL), Argonne, IL (United States). Argonne Leadership Computing Facility (ALCF)
University of Chicago, IL (United States)
National Institutes of Health (NIH), Bethesda, MD (United States)
Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)
Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM); University of Chicago, IL (United States)
University of Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)

Machine learning (ML) techniques with the genetic algorithm (GA) have been applied to determine a polarizable force field parameters using only ab initio data from quantum mechanics (QM) calculations of molecular clusters at the MP2/6-31G(d,p), DFMP2(fc)/jul-cc-pVDZ, and DFMP2(fc)/jul-cc-pVTZ levels to predict experimental condensed phase properties (i.e., density and heat of vaporization). The performance of this ML/GA approach is demonstrated on 4943 dimer electrostatic potentials and 1250 cluster interaction energies for methanol. Excellent agreement between the training data set from QM calculations and the optimized force field model was achieved. Our results were further improved by introducing an offset factor during the machine learning process to compensate for the discrepancy between the QM calculated energy and the energy reproduced by optimized force field, while maintaining the local “shape” of the QM energy surface. Throughout the machine learning process, experimental observables were not involved in the objective function, but were only used for model validation. The best model, optimized from the QM data at the DFMP2(fc)/jul-cc-pVTZ level, appears to perform even better than the original AMOEBA force field (amoeba09.prm), which was optimized empirically to match liquid properties. The present effort shows the possibility of using machine learning techniques to develop descriptive polarizable force field using only QM data. The ML/GA strategy to optimize force fields parameters described here could easily be extended to other molecular systems.

View Accepted Manuscript (DOE)

Cite

Export

Save

Research Organization:: Argonne National Laboratory (ANL), Argonne, IL (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Basic Energy Sciences (BES); University of Chicago Research Computing Center; National Institutes of Health (NIH)

Grant/Contract Number:: AC02-06CH11357

OSTI ID:: 1415482

Journal Information:: Journal of Chemical Theory and Computation, Journal Name: Journal of Chemical Theory and Computation Journal Issue: 9 Vol. 13; ISSN 1549-9618

Publisher:: American Chemical SocietyCopyright Statement

Country of Publication:: United States

Language:: English

References (69)

Psi4: an open-source ab initio electronic structure program : Psi4: an electronic structure program Turney, Justin M.; Simmonett, Andrew C.; Parrish, Robert M. Wiley Interdisciplinary Reviews: Computational Molecular Science, Vol. 2, Issue 4 https://doi.org/10.1002/wcms.93	journal	October 2011
Publisher’s Note: “Dispersion-corrected Møller–Plesset second-order perturbation theory” [J. Chem. Phys. 131, 094106 (2009)] Tkatchenko, Alexandre; DiStasio, Robert A.; Head-Gordon, Martin The Journal of Chemical Physics, Vol. 131, Issue 12 https://doi.org/10.1063/1.3240866	journal	September 2009
Monte carlo study of liquid water with two- and three-body ab initio potentials Clementi, E.; Corongiu, G. International Journal of Quantum Chemistry, Vol. 24, Issue S10 https://doi.org/10.1002/qua.560240707	journal	June 2009
Representation of Ion–Protein Interactions Using the Drude Polarizable Force-Field Li, Hui; Ngo, Van; Da Silva, Mauricio Chagas The Journal of Physical Chemistry B, Vol. 119, Issue 29 https://doi.org/10.1021/jp510560k	journal	September 2014
Structures of hydrogen bond networks formed by a few tens of methanol molecules in the gas phase: size-selective infrared spectroscopy of neutral and protonated methanol clusters Kobayashi, Tomohiro; Shishido, Ryunosuke; Mizuse, Kenta Physical Chemistry Chemical Physics, Vol. 15, Issue 24 https://doi.org/10.1039/c3cp50985k	journal	January 2013
Distributed multipole analysis Methods and applications Stone, A. J.; Alderton, M. Molecular Physics, Vol. 100, Issue 1 https://doi.org/10.1080/00268970110089432	journal	January 2002
Benchmark database of accurate (MP2 and CCSD(T) complete basis set limit) interaction energies of small model complexes, DNA base pairs, and amino acid pairs Jurečka, Petr; Šponer, Jiří; Černý, Jiří Physical Chemistry Chemical Physics, Vol. 8, Issue 17, p. 1985-1993 https://doi.org/10.1039/B600027D	journal	January 2006
Calculation of the reaction field due to off-center point multipoles Kong, Yong; Ponder, Jay W. The Journal of Chemical Physics, Vol. 107, Issue 2 https://doi.org/10.1063/1.474409	journal	July 1997
Use of a non-pair-additive intermolecular potential function to fit quantum-mechanical data on water molecule interactions Campbell, Edwin S.; Mezei, Mihaly The Journal of Chemical Physics, Vol. 67, Issue 5 https://doi.org/10.1063/1.435069	journal	September 1977
The origin of deficiency of the supermolecule second-order Møller-Plesset approach for evaluating interaction energies Cybulski, Slawomir M.; Lytle, Marion L. The Journal of Chemical Physics, Vol. 127, Issue 14 https://doi.org/10.1063/1.2795693	journal	October 2007
Accurate modeling of the intramolecular electrostatic energy of proteins Dudek, Michael J.; Ponder, Jay W. Journal of Computational Chemistry, Vol. 16, Issue 7 https://doi.org/10.1002/jcc.540160702	journal	July 1995
All-Atom Empirical Potential for Molecular Modeling and Dynamics Studies of Proteins ^† MacKerell, A. D.; Bashford, D.; Bellott, M. The Journal of Physical Chemistry B, Vol. 102, Issue 18 https://doi.org/10.1021/jp973084f	journal	April 1998
Transferable Next-Generation Force Fields from Simple Liquids to Complex Materials Schmidt, J. R.; Yu, Kuang; McDaniel, Jesse G. Accounts of Chemical Research, Vol. 48, Issue 3 https://doi.org/10.1021/ar500272n	journal	February 2015
Flexible, ab initio potential, and dipole moment surfaces for water. I. Tests and applications for clusters up to the 22-mer Wang, Yimin; Huang, Xinchuan; Shepler, Benjamin C. The Journal of Chemical Physics, Vol. 134, Issue 9 https://doi.org/10.1063/1.3554905	journal	March 2011
Determination of best-fit potential parameters for a reactive force field using a genetic algorithm Pahari, Poonam; Chaturvedi, Shashank Journal of Molecular Modeling, Vol. 18, Issue 3 https://doi.org/10.1007/s00894-011-1124-2	journal	June 2011
Distributed multipole analysis: Methods and applications Stone, A. J.; Alderton, M. Molecular Physics, Vol. 56, Issue 5 https://doi.org/10.1080/00268978500102891	journal	December 1985
The possibility of obtaining an effective pairwise additive intermolecular potential via an ab initio route by fitting to a cooperative model of condensed-phase configurations Mezei, Mihaly The Journal of Physical Chemistry, Vol. 95, Issue 18 https://doi.org/10.1021/j100171a059	journal	September 1991
Molecular modeling and dynamics studies with explicit inclusion of electronic polarizability: theory and applications Lopes, Pedro E. M.; Roux, Benoit; MacKerell, Alexander D. Theoretical Chemistry Accounts, Vol. 124, Issue 1-2 https://doi.org/10.1007/s00214-009-0617-x	journal	August 2009
Distributed Multipole Analysis: Stability for Large Basis Sets Stone, Anthony J. Journal of Chemical Theory and Computation, Vol. 1, Issue 6 https://doi.org/10.1021/ct050190+	journal	September 2005
A simple polarizable model of water based on classical Drude oscillators Lamoureux, Guillaume; MacKerell, Alexander D.; Roux, Benoı̂t The Journal of Chemical Physics, Vol. 119, Issue 10 https://doi.org/10.1063/1.1598191	journal	September 2003
Steepest descent methods for multicriteria optimization Fliege, Jörg; Svaiter, Benar Fux Mathematical Methods of Operations Research (ZOR), Vol. 51, Issue 3 https://doi.org/10.1007/s001860000043	journal	August 2000
CI study of the water dimer potential surface Matsuoka, O.; Clementi, E.; Yoshimine, M. The Journal of Chemical Physics, Vol. 64, Issue 4 https://doi.org/10.1063/1.432402	journal	February 1976
eReaxFF: A Pseudoclassical Treatment of Explicit Electrons within Reactive Force Field Simulations Islam, Md Mahbubul; Kolesov, Grigory; Verstraelen, Toon Journal of Chemical Theory and Computation, Vol. 12, Issue 8 https://doi.org/10.1021/acs.jctc.6b00432	journal	July 2016
First-Principles Many-Body Force Fields from the Gas Phase to Liquid: A “Universal” Approach McDaniel, Jesse G.; Schmidt, J. R. The Journal of Physical Chemistry B, Vol. 118, Issue 28 https://doi.org/10.1021/jp501128w	journal	April 2014
An efficient newton-like method for molecular mechanics energy minimization of large molecules Ponder, Jay W.; Richards, Frederic M. Journal of Computational Chemistry, Vol. 8, Issue 7 https://doi.org/10.1002/jcc.540080710	journal	October 1987
Modeling induced polarization with classical Drude oscillators: Theory and molecular dynamics simulation algorithm Lamoureux, Guillaume; Roux, Benoı̂t The Journal of Chemical Physics, Vol. 119, Issue 6 https://doi.org/10.1063/1.1589749	journal	August 2003
Application of Gaussian Electrostatic Model (GEM) Distributed Multipoles in the AMOEBA Force Field Cisneros, G. Andrés Journal of Chemical Theory and Computation, Vol. 8, Issue 12 https://doi.org/10.1021/ct300630u	journal	September 2012
Ab Initio -Based Bond Order Potential to Investigate Low Thermal Conductivity of Stanene Nanostructures Cherukara, Mathew J.; Narayanan, Badri; Kinaci, Alper The Journal of Physical Chemistry Letters, Vol. 7, Issue 19 https://doi.org/10.1021/acs.jpclett.6b01562	journal	September 2016
Non-additivity in water-ion-water interactions Clementi, Enrico; Kistenmacher, Hans; Kołos, Włodzimierz Theoretica Chimica Acta, Vol. 55, Issue 4 https://doi.org/10.1007/BF00549424	journal	December 1980
Roulette-wheel selection via stochastic acceptance Lipowski, Adam; Lipowska, Dorota Physica A: Statistical Mechanics and its Applications, Vol. 391, Issue 6 https://doi.org/10.1016/j.physa.2011.12.004	journal	March 2012
An Empirical Polarizable Force Field Based on the Classical Drude Oscillator Model: Development History and Recent Applications Lemkul, Justin A.; Huang, Jing; Roux, Benoît Chemical Reviews, Vol. 116, Issue 9 https://doi.org/10.1021/acs.chemrev.5b00505	journal	January 2016
Accurate modeling of intermolecular forces: a systematic Møller-Plesset study of the argon dimer using correlation consistent basis sets Woon, David E. Chemical Physics Letters, Vol. 204, Issue 1-2 https://doi.org/10.1016/0009-2614(93)85601-J	journal	March 1993
The Structure of Electronic Excitation Levels in Insulating Crystals Wannier, Gregory H. Physical Review, Vol. 52, Issue 3 https://doi.org/10.1103/PhysRev.52.191	journal	August 1937
Genetic algorithms for modelling and optimisation McCall, John Journal of Computational and Applied Mathematics, Vol. 184, Issue 1 https://doi.org/10.1016/j.cam.2004.07.034	journal	December 2005
Getting the Right Answers for the Right Reasons: Toward Predictive Molecular Simulations of Water with Many-Body Potential Energy Functions Paesani, Francesco Accounts of Chemical Research, Vol. 49, Issue 9 https://doi.org/10.1021/acs.accounts.6b00285	journal	August 2016
Nonconvex minimization calculations and the conjugate gradient method Powell, M. J. D. Numerical Analysis https://doi.org/10.1007/BFb0099521	book	January 1984
The structure of liquid methanol revisited: a neutron diffraction experiment at −80 °C and +25 °C Yamaguchi, T.; Hidaka, K.; Soper, A. K. Molecular Physics, Vol. 96, Issue 8 https://doi.org/10.1080/00268979909483060	journal	April 1999
Scaled MP3 Non-Covalent Interaction Energies Agree Closely with Accurate CCSD(T) Benchmark Data Pitoňák, Michal; Neogrády, Pavel; Černý, Jiří ChemPhysChem, Vol. 10, Issue 1 https://doi.org/10.1002/cphc.200800718	journal	January 2009
Distributed multipole analysis, or how to describe a molecular charge distribution Stone, A. J. Chemical Physics Letters, Vol. 83, Issue 2 https://doi.org/10.1016/0009-2614(81)85452-8	journal	October 1981
A new water potential including polarization: Application to gas‐phase, liquid, and crystal properties of water Cieplak, Piotr; Kollman, Peter; Lybrand, Terry The Journal of Chemical Physics, Vol. 92, Issue 11 https://doi.org/10.1063/1.458260	journal	June 1990
Automation of AMOEBA polarizable force field parameterization for small molecules Wu, Johnny C.; Chattree, Gaurav; Ren, Pengyu Theoretical Chemistry Accounts, Vol. 131, Issue 3 https://doi.org/10.1007/s00214-012-1138-6	journal	February 2012
Ab initio calculation of electrostatic multipoles with Wannier functions for large-scale biomolecular simulations Sagui, Celeste; Pomorski, Pawel; Darden, Thomas A. The Journal of Chemical Physics, Vol. 120, Issue 9 https://doi.org/10.1063/1.1644800	journal	March 2004
Polarizable Atomic Multipole-Based Molecular Mechanics for Organic Molecules Ren, Pengyu; Wu, Chuanjie; Ponder, Jay W. Journal of Chemical Theory and Computation, Vol. 7, Issue 10 https://doi.org/10.1021/ct200304d	journal	August 2011
Recent applications and developments of charge equilibration force fields for modeling dynamical charges in classical molecular dynamics simulations Bauer, Brad A.; Patel, Sandeep Theoretical Chemistry Accounts, Vol. 131, Issue 3 https://doi.org/10.1007/s00214-012-1153-7	journal	March 2012
Self-consistent molecular orbital methods. 21. Small split-valence basis sets for first-row elements Binkley, J. Stephen; Pople, John A.; Hehre, Warren J. Journal of the American Chemical Society, Vol. 102, Issue 3 https://doi.org/10.1021/ja00523a008	journal	January 1980
Strike a balance: Optimization of backbone torsion parameters of AMBER polarizable force field for simulations of proteins and peptides Wang, Zhi-Xiang; Zhang, Wei; Wu, Chun Journal of Computational Chemistry, Vol. 27, Issue 6 https://doi.org/10.1002/jcc.20386	journal	January 2006
Global optimization of parameters in the reactive force field ReaxFF for SiOH Larsson, Henrik R.; van Duin, Adri C. T.; Hartke, Bernd Journal of Computational Chemistry, Vol. 34, Issue 25 https://doi.org/10.1002/jcc.23382	journal	July 2013
Current Status of the AMOEBA Polarizable Force Field Ponder, Jay W.; Wu, Chuanjie; Ren, Pengyu The Journal of Physical Chemistry B, Vol. 114, Issue 8 https://doi.org/10.1021/jp910674d	journal	March 2010
The calculation of small molecular interactions by the differences of separate total energies. Some procedures with reduced errors Boys, S. F.; Bernardi, F. Molecular Physics, Vol. 100, Issue 1 https://doi.org/10.1080/00268970110088901	journal	January 2002
Describing Noncovalent Interactions beyond the Common Approximations: How Accurate Is the “Gold Standard,” CCSD(T) at the Complete Basis Set Limit? Řezáč, Jan; Hobza, Pavel Journal of Chemical Theory and Computation, Vol. 9, Issue 5 https://doi.org/10.1021/ct400057w	journal	April 2013
Polarizable Atomic Multipole Water Model for Molecular Mechanics Simulation Ren, Pengyu; Ponder, Jay W. The Journal of Physical Chemistry B, Vol. 107, Issue 24 https://doi.org/10.1021/jp027815+	journal	May 2003
Towards accurate prediction of catalytic activity in IrO ₂ nanoclusters via first principles-based variable charge force field Sen, F. G.; Kinaci, A.; Narayanan, B. Journal of Materials Chemistry A, Vol. 3, Issue 37 https://doi.org/10.1039/C5TA04678E	journal	January 2015
ReaxFF: A Reactive Force Field for Hydrocarbons van Duin, Adri C. T.; Dasgupta, Siddharth; Lorant, Francois The Journal of Physical Chemistry A, Vol. 105, Issue 41 https://doi.org/10.1021/jp004368u	journal	October 2001
Physically-Motivated Force Fields from Symmetry-Adapted Perturbation Theory McDaniel, Jesse G.; Schmidt, J. R. The Journal of Physical Chemistry A, Vol. 117, Issue 10 https://doi.org/10.1021/jp3108182	journal	February 2013
The representation of van der Waals (vdW) interactions in molecular mechanics force fields: potential form, combination rules, and vdW parameters Halgren, Thomas A. Journal of the American Chemical Society, Vol. 114, Issue 20 https://doi.org/10.1021/ja00046a032	journal	September 1992
Effective force fields for condensed phase systems from ab initio molecular dynamics simulation: A new method for force-matching Izvekov, Sergei; Parrinello, Michele; Burnham, Christian J. The Journal of Chemical Physics, Vol. 120, Issue 23 https://doi.org/10.1063/1.1739396	journal	June 2004
Maximally localized generalized Wannier functions for composite energy bands Marzari, Nicola; Vanderbilt, David Physical Review B, Vol. 56, Issue 20 https://doi.org/10.1103/PhysRevB.56.12847	journal	November 1997
Empirical force fields for biological macromolecules: Overview and issues Mackerell, Alexander D. Journal of Computational Chemistry, Vol. 25, Issue 13 https://doi.org/10.1002/jcc.20082	journal	January 2004
Algorithms for calculating excluded volume and its derivatives as a function of molecular conformation and their use in energy minimization Kundrot, Craig E.; Ponder, Jay W.; Richards, Frederic M. Journal of Computational Chemistry, Vol. 12, Issue 3 https://doi.org/10.1002/jcc.540120314	journal	April 1991
Levels of symmetry adapted perturbation theory (SAPT). I. Efficiency and performance for interaction energies Parker, Trent M.; Burns, Lori A.; Parrish, Robert M. The Journal of Chemical Physics, Vol. 140, Issue 9 https://doi.org/10.1063/1.4867135	journal	March 2014
Analysis and Application of Potential Energy Smoothing and Search Methods for Global Optimization Pappu, Rohit V.; Hart, Reece K.; Ponder, Jay W. The Journal of Physical Chemistry B, Vol. 102, Issue 48 https://doi.org/10.1021/jp982255t	journal	November 1998
Predictions of the Properties of Water from First Principles Bukowski, R.; Szalewicz, K.; Groenenboom, G. C. Science, Vol. 315, Issue 5816 https://doi.org/10.1126/science.1136371	journal	March 2007
Gaussian basis sets for use in correlated molecular calculations. I. The atoms boron through neon and hydrogen Dunning, Thom H. The Journal of Chemical Physics, Vol. 90, Issue 2 https://doi.org/10.1063/1.456153	journal	January 1989
MP2 energy evaluation by direct methods Head-Gordon, Martin; Pople, John A.; Frisch, Michael J. Chemical Physics Letters, Vol. 153, Issue 6 https://doi.org/10.1016/0009-2614(88)85250-3	journal	December 1988
Next-Generation Force Fields from Symmetry-Adapted Perturbation Theory McDaniel, Jesse G.; Schmidt, J. R. Annual Review of Physical Chemistry, Vol. 67, Issue 1 https://doi.org/10.1146/annurev-physchem-040215-112047	journal	May 2016
Development of a Modified Embedded Atom Force Field for Zirconium Nitride Using Multi-Objective Evolutionary Optimization Narayanan, Badri; Sasikumar, Kiran; Mei, Zhi-Gang The Journal of Physical Chemistry C, Vol. 120, Issue 31 https://doi.org/10.1021/acs.jpcc.6b05296	journal	July 2016
Dynamical fluctuating charge force fields: Application to liquid water Rick, Steven W.; Stuart, Steven J.; Berne, B. J. The Journal of Chemical Physics, Vol. 101, Issue 7 https://doi.org/10.1063/1.468398	journal	October 1994
Calculations of electrostatic properties in proteins Van Belle, Daniel; Couplet, Ignace; Prevost, Martine Journal of Molecular Biology, Vol. 198, Issue 4 https://doi.org/10.1016/0022-2836(87)90213-0	journal	December 1987
Describing the Diverse Geometries of Gold from Nanoclusters to Bulk—A First-Principles-Based Hybrid Bond-Order Potential Narayanan, Badri; Kinaci, Alper; Sen, Fatih G. The Journal of Physical Chemistry C, Vol. 120, Issue 25 https://doi.org/10.1021/acs.jpcc.6b02934	journal	June 2016

Cited By (1)

The TensorMol-0.1 model chemistry: a neural network augmented with long-range physics Yao, Kun; Herr, John E.; Toth, David W. Chemical Science, Vol. 9, Issue 8 https://doi.org/10.1039/c7sc04934j	journal	January 2018

Similar Records

Density functional theory study of the interaction of vinyl radical, ethyne, and ethene with benzene, aimed to define an affordable computational level to investigate stability trends in large van der Waals complexes

Journal Article · 2013 · Journal of Chemical Physics · OSTI ID:22253231

Maranzana, Andrea; Giordana, Anna; Indarto, Antonius; +4 more

Atlas of putative minima and low-lying energy networks of water clusters n = 3 - 25

Journal Article · 2019 · Journal of Chemical Physics · OSTI ID:1580513

Rakshit, Avijit; Bandyopadhyay, Pradipta; Heindel, Joseph; +1 more

Dual-level reaction-path dynamics (the III approach to VTST with semiclassical tunneling). Application to OH + NH[sub 3] [yields] H[sub 2]O + NH[sub 2]

Journal Article · 1995 · Journal of Physical Chemistry; (United States) · OSTI ID:6582737

Corchado, J C; Espinosa-Garcia, J; Hu, W P; +2 more

Related Subjects

97
force field
machine learning
ab initio data

Title: Machine Learning Force Field Parameters from Ab Initio Data

Citation Formats

References (69)

Cited By (1)

Similar Records

Related Subjects