Systematic Parametrization of Polarizable Force Fields from Quantum Chemistry Data

Wang, Lee‐Ping; Chen, Jiahao; Voorhis, Troy Van

doi:10.1021/ct300826t

Cited by 216 publications

(301 citation statements)

References 90 publications

Supporting

Mentioning

299

Contrasting

Order By: Relevance

“…The second vector s⃗2, that is, the vector along which the data show a significant variation, numerically corresponds to a normalized vector u⃗2 = (nH −nX), also determined by the stoichiometry. Thus, the eigenbasis of the covariance matrix Σ̃ can be represented as (21) The dramatic difference in the data variation along the two covariance eigenvectors suggests that the fitness function has very different curvatures along these two directions. This curvature of the fitness function can be examined explicitly by computing and diagonalizing its Hessian matrix or, for simplicity, the Hessian of the LS sum H (eq 5): 93 (22) (23) As can be seen from Figure 5 and Table S3 in the Supporting Information, the Hessian eigenbases H̃ computed for all four twocharge molecules are numerically identical to the corresponding covariance matrix eigenbases Σ̃ and the basis of normalized vectors: u⃗i, Ũ: There is an inverse relationship between the eigenvalues of the fitness function/LS sum Hessian and the covariance matrices: the Hessian eigenvector h⃗2 with near-zero eigenvalue/curvature corresponds to the covariance eigenvector s⃗2 with a large variance; at the same time, the Hessian eigenvector h⃗1 with a large curvature corresponds to the covariance eigenvector s⃗1 with near-zero variance.…”

Section: Covariance Matrix Analysis Of Ga Resultsmentioning

confidence: 99%

“…However, simultaneous fitting of several parameters describing intermolecular interactions (point charges, Lennard-Jones parameters, and in the case of polarizable force fields, atomic polarizabilities) may significantly improve the accuracy of force field description. 20,21 These simultaneous optimizations of different force field terms can take advantage of extensive training sets that can be easily generated using electronic structure calculations and may include data on the intermolecular interaction energies. [22][23][24][25][26] Moreover, in this approach the fitted interaction energy would implicitly include the polarization effects, even staying within the fixed point-charge force field framework.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Genetic Algorithm Optimization of Point Charges in Force Field Development: Challenges and Insights

2015

View full text Add to dashboard Cite

Evolutionary methods, such as genetic algorithms (GAs), provide powerful tools for optimization of the force field parameters, especially in the case of simultaneous fitting of the force field terms against extensive reference data. However, GA fitting of the nonbonded interaction parameters that includes point charges has not been explored in the literature, likely due to numerous difficulties with even a simpler problem of the least-squares fitting of the atomic point charges against a reference molecular electrostatic potential (MEP), which often demonstrates an unusually high variation of the fitted charges on buried atoms. Here, we examine the performance of the GA approach for the least-squares MEP point charge fitting, and show that the GA optimizations suffer from a magnified version of the classical buried atom effect, producing highly scattered yet correlated solutions. This effect can be understood in terms of the linearly independent, natural coordinates of the MEP fitting problem defined by the eigenvectors of the least-squares sum Hessian matrix, which are also equivalent to the eigenvectors of the covariance matrix evaluated for the scattered GA solutions. GAs quickly converge with respect to the high-curvature coordinates defined by the eigenvectors related to the leading terms of the multipole expansion, but have difficulty converging with respect to the low-curvature coordinates that mostly depend on the buried atom charges. The performance of the evolutionary techniques dramatically improves when the point charge optimization is performed using the Hessian or covariance matrix eigenvectors, an approach with a significant potential for the evolutionary optimization of the fixed-charge biomolecular force fields.

show abstract

Section: Covariance Matrix Analysis Of Ga Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Genetic Algorithm Optimization of Point Charges in Force Field Development: Challenges and Insights

2015

View full text Add to dashboard Cite

show abstract

“…Here, all the AMOEBA solvent models were prepared using the parameters taken from amoeba09.prm39 except for chloroform 42. The most recent AMOEBA chloroform parameters published by Ren and coworkers,42 which made use of the ForceBalance parameter optimization protocol, were used 44. For fixed‐charge simulations, solvent parameters were taken from Cieplak et al (chloroform),45 Grabuleda et al (actetonitrile),46 and Dupradeau et al (toluene and DMSO) 47.…”

Section: Methodsmentioning

confidence: 99%

Evaluation of solvation free energies for small molecules with the AMOEBA polarizable force field

2016

View full text Add to dashboard Cite

The effects of electronic polarization in biomolecular interactions will differ depending on the local dielectric constant of the environment, such as in solvent, DNA, proteins, and membranes. Here the performance of the AMOEBA polarizable force field is evaluated under nonaqueous conditions by calculating the solvation free energies of small molecules in four common organic solvents. Results are compared with experimental data and equivalent simulations performed with the GAFF pairwise‐additive force field. Although AMOEBA results give mean errors close to “chemical accuracy,” GAFF performs surprisingly well, with statistically significantly more accurate results than AMOEBA in some solvents. However, for both models, free energies calculated in chloroform show worst agreement to experiment and individual solutes are consistently poor performers, suggesting non‐potential‐specific errors also contribute to inaccuracy. Scope for the improvement of both potentials remains limited by the lack of high quality experimental data across multiple solvents, particularly those of high dielectric constant. © 2016 The Authors. Journal of Computational Chemistry Published by Wiley Periodicals, Inc.

show abstract

“…Working in this basis allows direct connection to experiment and often provides insight into the molecular interactions driving biophysical phenomena. For example, the projection onto observables could be used to rationally infer force field parameters-essentially a Bayesian version of the ForceBalance method (54,55). Third, BELT does not require subjective choices.…”

mentioning

confidence: 99%

Bayesian Energy Landscape Tilting: Towards Concordant Models of Molecular Ensembles

Beauchamp¹,

Pande²,

Das³

2014

Preprint

View full text Add to dashboard Cite

Predicting biological structure has remained challenging for systems such as disordered proteins that take on myriad conformations. Hybrid simulation/experiment strategies have been undermined by difficulties in evaluating errors from computational model inaccuracies and data uncertainties. Building on recent proposals from maximum entropy theory and nonequilibrium thermodynamics, we address these issues through a Bayesian Energy Landscape Tilting (BELT) scheme for computing Bayesian "hyperensembles" over conformational ensembles. BELT uses Markov chain Monte Carlo to directly sample maximum-entropy conformational ensembles consistent with a set of input experimental observables. To test this framework, we apply BELT to model trialanine, starting from disagreeing simulations with the force fields ff96, ff99, ff99sbnmr-ildn, CHARMM27, and OPLS-AA. BELT incorporation of limited chemical shift and 3 J measurements gives convergent values of the peptide's α, β, and P P II conformational populations in all cases. As a test of predictive power, all five BELT hyperensembles recover set-aside measurements not used in the fitting and report accurate errors, even when starting from highly inaccurate simulations. BELT's principled framework thus enables practical predictions for complex biomolecular systems from discordant simulations and sparse data.

show abstract

Systematic Parametrization of Polarizable Force Fields from Quantum Chemistry Data

Cited by 216 publications

References 90 publications

Genetic Algorithm Optimization of Point Charges in Force Field Development: Challenges and Insights

Genetic Algorithm Optimization of Point Charges in Force Field Development: Challenges and Insights

Evaluation of solvation free energies for small molecules with the AMOEBA polarizable force field

Bayesian Energy Landscape Tilting: Towards Concordant Models of Molecular Ensembles

Contact Info

Product

Resources

About