Analyzing Multimodal Probability Measures with Autoencoders.

Finding collective variables to describe some important coarse-grained information on physical systems, in particular metastable states, remains a key issue in molecular dynamics. Recently, machine learning techniques have been intensively used to complement and possibly bypass expert knowledge in order to construct collective variables. Our focus here is on neural network approaches based on autoencoders. We study some relevant mathematical properties of the loss function considered for training autoencoders and provide physical interpretations based on conditional variances and minimum energy paths. We also consider various extensions in order to better describe physical systems, by incorporating more information on transition states at saddle points, and/or allowing for multiple decoders in order to describe several transition paths. Our results are illustrated on toy two-dimensional systems and on alanine dipeptide.

Autoencoders for dimensionality reduction in molecular dynamics: Collective variable dimension, biasing, and transition states.

The heat shock protein 90 (Hsp90) is a molecular chaperone that controls the folding and activation of client proteins using the free energy of ATP hydrolysis. The Hsp90 active site is in its N-terminal domain (NTD). Our goal is to characterize the dynamics of NTD using an autoencoder-learned collective variable (CV) in conjunction with adaptive biasing force Langevin dynamics. Using dihedral analysis, we cluster all available experimental Hsp90 NTD structures into distinct native states. We then perform unbiased molecular dynamics (MD) simulations to construct a dataset that represents each state and use this dataset to train an autoencoder. Two autoencoder architectures are considered, with one and two hidden layers, respectively, and bottlenecks of dimension k ranging from 1 to 10. We demonstrate that the addition of an extra hidden layer does not significantly improve the performance, while it leads to complicated CVs that increase the computational cost of biased MD calculations. In addition, a two-dimensional (2D) bottleneck can provide enough information of the different states, while the optimal bottleneck dimension is five. For the 2D bottleneck, the 2D CV is directly used in biased MD simulations. For the five-dimensional (5D) bottleneck, we perform an analysis of the latent CV space and identify the pair of CV coordinates that best separates the states of Hsp90. Interestingly, selecting a 2D CV out of the 5D CV space leads to better results than directly learning a 2D CV and allows observation of transitions between native states when running free energy biased dynamics.

Computing Surface Reaction Rates by Adaptive Multilevel Splitting Combined with Machine Learning and Ab Initio Molecular Dynamics.

Computing accurate rate constants for catalytic events occurring at the surface of a given material represents a challenging task with multiple potential applications in chemistry. To address this question, we propose an approach based on a combination of the rare event sampling method called adaptive multilevel splitting (AMS) and ab initio molecular dynamics. The AMS method requires a one-dimensional reaction coordinate to index the progress of the transition. Identifying a good reaction coordinate is difficult, especially for high dimensional problems such as those encountered in catalysis. We probe various approaches to build reaction coordinates such as support vector machine and path collective variables. The AMS is implemented so as to communicate with a density functional theory-plane wave code. A relevant case study in catalysis, the change of conformation and the dissociation of a water molecule chemisorbed on the (100) γ-alumina surface, is used to evaluate our approach. The calculated rate constants and transition mechanisms are discussed and compared to those obtained by a conventional static approach based on the Eyring-Polanyi equation with harmonic approximation. It is revealed that the AMS method may provide rate constants that are smaller than those provided by the static approach by up to 2 orders of magnitude due to entropic effects involved in the chemisorbed water molecule.

Frequency and field-dependent response of confined electrolytes from Brownian dynamics simulations.

Using Brownian dynamics simulations, we investigate the effects of confinement, adsorption on surfaces, and ion-ion interactions on the response of confined electrolyte solutions to oscillating electric fields in the direction perpendicular to the confining walls. Nonequilibrium simulations allows to characterize the transitions between linear and nonlinear regimes when varying the magnitude and frequency of the applied field, but the linear response, characterized by the frequency-dependent conductivity, is more efficiently predicted from the equilibrium current fluctuations. To that end, we (rederive and) use the Green-Kubo relation appropriate for overdamped dynamics, which differs from the standard one for Newtonian or underdamped Langevin dynamics. This expression highlights the contributions of the underlying Brownian fluctuations and of the interactions of the particles between them and with external potentials. Although already known in the literature, this relation has rarely been used to date, beyond the static limit to determine the effective diffusion coefficient or the DC conductivity. The frequency-dependent conductivity always decays from a bulk-like behavior at high frequency to a vanishing conductivity at low frequency due to the confinement of the charge carriers by the walls. We discuss the characteristic features of the crossover between the two regimes, most importantly how the crossover frequency depends on the confining distance and the salt concentration, and the fact that adsorption on the walls may lead to significant changes both at high and low frequencies. Conversely, our results illustrate the possibility to obtain information on diffusion between walls, charge relaxation, and adsorption by analyzing the frequency-dependent conductivity.

Chasing Collective Variables Using Autoencoders and Biased Trajectories.

Free energy biasing methods have proven to be powerful tools to accelerate the simulation of important conformational changes of molecules by modifying the sampling measure. However, most of these methods rely on the prior knowledge of low-dimensional slow degrees of freedom, i.e., collective variables (CVs). Alternatively, such CVs can be identified using machine learning (ML) and dimensionality reduction algorithms. In this context, approaches where the CVs are learned in an iterative way using adaptive biasing have been proposed: at each iteration, the learned CV is used to perform free energy adaptive biasing to generate new data and learn a new CV. In this paper, we introduce a new iterative method involving CV learning with autoencoders: Free Energy Biasing and Iterative Learning with AutoEncoders (FEBILAE). Our method includes a reweighting scheme to ensure that the learning model optimizes the same loss at each iteration and achieves CV convergence. Using the alanine dipeptide system and the solvated chignolin mini-protein system as examples, we present results of our algorithm using the extended adaptive biasing force as the free energy adaptive biasing method.

Machine Learning Force Fields and Coarse-Grained Variables in Molecular Dynamics: Application to Materials and Biological Systems.

Machine learning encompasses tools and algorithms that are now becoming popular in almost all scientific and technological fields. This is true for molecular dynamics as well, where machine learning offers promises of extracting valuable information from the enormous amounts of data generated by simulation of complex systems. We provide here a review of our current understanding of goals, benefits, and limitations of machine learning techniques for computational studies on atomistic systems, focusing on the construction of empirical force fields from ab initio databases and the determination of reaction coordinates for free energy computation and enhanced sampling.

Smoothed Biasing Forces Yield Unbiased Free Energies with the Extended-System Adaptive Biasing Force Method.

We report a theoretical description and numerical tests of the extended-system adaptive biasing force method (eABF), together with an unbiased estimator of the free energy surface from eABF dynamics. Whereas the original ABF approach uses its running estimate of the free energy gradient as the adaptive biasing force, eABF is built on the idea that the exact free energy gradient is not necessary for efficient exploration, and that it is still possible to recover the exact free energy separately with an appropriate estimator. eABF does not directly bias the collective coordinates of interest, but rather fictitious variables that are harmonically coupled to them; therefore is does not require second derivative estimates, making it easily applicable to a wider range of problems than ABF. Furthermore, the extended variables present a smoother, coarse-grain-like sampling problem on a mollified free energy surface, leading to faster exploration and convergence. We also introduce CZAR, a simple, unbiased free energy estimator from eABF trajectories. eABF/CZAR converges to the physical free energy surface faster than standard ABF for a wide range of parameters.

Size consistency in smoothed dissipative particle dynamics.

Smoothed dissipative particle dynamics (SDPD) is a mesoscopic method that allows one to select the level of resolution at which a fluid is simulated. In this work, we study the consistency of the resulting thermodynamic properties as a function of the size of the mesoparticles, both at equilibrium and out of equilibrium. We also propose a reformulation of the SDPD equations in terms of energy variables. This increases the similarities with dissipative particle dynamics with energy conservation and opens the way for a coupling between the two methods. Finally, we present a numerical scheme for SDPD that ensures the conservation of the invariants of the dynamics. Numerical simulations illustrate this approach.

New parallelizable schemes for integrating the Dissipative Particle Dynamics with Energy conservation.

This work presents new parallelizable numerical schemes for the integration of dissipative particle dynamics with energy conservation. So far, no numerical scheme introduced in the literature is able to correctly preserve the energy over long times and give rise to small errors on average properties for moderately small time steps, while being straightforwardly parallelizable. We present in this article two new methods, both straightforwardly parallelizable, allowing to correctly preserve the total energy of the system. We illustrate the accuracy and performance of these new schemes both on equilibrium and nonequilibrium parallel simulations.

Permutation-invariant distance between atomic configurations.

We present a permutation-invariant distance between atomic configurations, defined through a functional representation of atomic positions. This distance enables us to directly compare different atomic environments with an arbitrary number of particles, without going through a space of reduced dimensionality (i.e., fingerprints) as an intermediate step. Moreover, this distance is naturally invariant through permutations of atoms, avoiding the time consuming associated minimization required by other common criteria (like the root mean square distance). Finally, the invariance through global rotations is accounted for by a minimization procedure in the space of rotations solved by Monte Carlo simulated annealing. A formal framework is also introduced, showing that the distance we propose verifies the property of a metric on the space of atomic configurations. Two examples of applications are proposed. The first one consists in evaluating faithfulness of some fingerprints (or descriptors), i.e., their capacity to represent the structural information of a configuration. The second application concerns structural analysis, where our distance proves to be efficient in discriminating different local structures and even classifying their degree of similarity.

Free energy calculations from adaptive molecular dynamics simulations with adiabatic reweighting.

We propose an adiabatic reweighting algorithm for computing the free energy along an external parameter from adaptive molecular dynamics simulations. The adaptive bias is estimated using Bayes identity and information from all the sampled configurations. We apply the algorithm to a structural transition in a cluster and to the migration of a crystalline defect along a reaction coordinate. Compared to standard adaptive molecular dynamics, we observe an acceleration of convergence. With the aid of the algorithm, it is also possible to iteratively construct the free energy along the reaction coordinate without having to differentiate the gradient of the reaction coordinate or any biasing potential.

Local density dependent potential for compressible mesoparticles.

This work proposes a coarse grained description of molecular systems based on mesoparticles representing several molecules, where interactions between mesoparticles are modelled by an interparticle potential of molecular type. Since strong non-equilibrium situations over a wide range of pressure and density are targeted, the internal compressibility of the mesoparticles has to be considered. This is done by introducing a dependence of the potential on the local environment of the mesoparticles. To define local densities, we resort to a three-dimensional Voronoi tessellation instead of standard local, spherical averages. As an example, a local density dependent potential is fitted to reproduce the Hugoniot curve of a model of nitromethane over a large range of pressures and densities.

Microscopic calculations of Hugoniot curves of neat triaminotrinitrobenzene (TATB) and of its detonation products.

We compute the Hugoniot curves of both neat triaminotrinitrobenzene (TATB) and its detonation products mixture using atomistic simulation tools. To compute the Hugoniot states, we adapted our sampling constraints in average (SCA) method (Maillet et al. Appl. Math. Res. eXpress 2009, 2008, abn004) to Monte Carlo simulations. For neat TATB, we show that the potential proposed by Rai (Rai et al. J. Chem. Phys. 2008, 129, 194510) is not accurate enough to predict the Hugoniot curve and requires some optimization of its parameters. Concerning the detonation products, thermodynamic properties at chemical equilibrium are computed using a specific reaction ensemble Monte Carlo (RxMC) method (Bourasseau et al. Phys. Chem. Chem. Phys. 2011, 13, 7060), taking into account the presence of carbon clusters in the fluid mixture. We show that this explicit description of the solid phase immersed in the fluid phase modifies the chemical equilibrium.

Negative thermal conductivity of chains of rotors with mechanical forcing.

We consider chains of rotors subjected to both thermal and mechanical forcings in a nonequilibrium steady state. Unusual nonlinear profiles of temperature and velocities are observed in the system. In particular, the temperature is maximal in the center, which is an indication of the nonlocal behavior of the system. Despite this uncommon behavior, local equilibrium holds for long enough chains. Our numerical results also show that when the mechanical forcing is strong enough, the energy current can be increased by an inverse temperature gradient. This counterintuitive result again reveals the complexity of nonequilibrium states.

Free energy calculations: an efficient adaptive biasing potential method.

We develop an efficient sampling and free energy calculation technique within the adaptive biasing potential (ABP) framework. By mollifying the density of states we obtain an approximate free energy and an adaptive bias potential that is computed directly from the population along the coordinates of the free energy. Because of the mollifier, the bias potential is "nonlocal", and its gradient admits a simple analytic expression. A single observation of the reaction coordinate can thus be used to update the approximate free energy at every point within a neighborhood of the observation. This greatly reduces the equilibration time of the adaptive bias potential. This approximation introduces two parameters: strength of mollification and the zero of energy of the bias potential. While we observe that the approximate free energy is a very good estimate of the actual free energy for a large range of mollification strength, we demonstrate that the errors associated with the mollification may be removed via deconvolution. The zero of energy of the bias potential, which is easy to choose, influences the speed of convergence but not the limiting accuracy. This method is simple to apply to free energy or mean force computation in multiple dimensions and does not involve second derivatives of the reaction coordinates, matrix manipulations nor on-the-fly adaptation of parameters. For the alanine dipeptide test case, the new method is found to gain as much as a factor of 10 in efficiency as compared to two basic implementations of the adaptive biasing force methods, and it is shown to be as efficient as well-tempered metadynamics with the postprocess deconvolution giving a clear advantage to the mollified density of states method.

Constant entropy sampling and release waves of shock compressions.

We present or recall several equilibrium methods that allow one to compute isentropic processes, either during the compression or the release of the material. These methods are applied to compute the isentropic release of a shocked monoatomic liquid at high pressure and temperature. Moreover, equilibrium results of isentropic release are compared to the direct nonequilibrium simulation of the same process. We show that due to the viscosity of the liquid but also to nonequilibrium effects, the release of the system is not strictly isentropic.

Many-body green function of degenerate systems.

A rigorous nonperturbative adiabatic approximation of the evolution operator in the many-body physics of degenerate systems is derived. This approximation is used to solve the long-standing problem of the choice of the initial states of H(0) leading to eigenstates of H(0) + V for degenerate systems. These initial states are eigenstates of P(0)VP(0), where P(0) is the projection onto a degenerate eigenspace of H(0). This result is used to give the proper definition of the Green function, the statistical Green function and the nonequilibrium Green function of degenerate systems. The convergence of these Green functions is established.

Computation of free energy profiles with parallel adaptive dynamics.

We propose a formulation of an adaptive computation of free energy differences, in the adaptive biasing force or nonequilibrium metadynamics spirit, using conditional distributions of samples of configurations which evolve in time. This allows us to present a truly unifying framework for these methods, and to prove convergence results for certain classes of algorithms. From a numerical viewpoint, a parallel implementation of these methods is very natural, the replicas interacting through the reconstructed free energy. We demonstrate how to improve this parallel implementation by resorting to some selection mechanism on the replicas. This is illustrated by computations on a model system of conformational changes.

The effective local potential method: implementation for molecules and relation to approximate optimized effective potential techniques.

We have recently formulated a new approach, named the effective local potential (ELP) method, for calculating local exchange-correlation potentials for orbital-dependent functionals based on minimizing the variance of the difference between a given nonlocal potential and its desired local counterpart [V. N. Staroverov et al., J. Chem. Phys. 125, 081104 (2006)]. Here we show that under a mildly simplifying assumption of frozen molecular orbitals, the equation defining the ELP has a unique analytic solution which is identical with the expression arising in the localized Hartree-Fock (LHF) and common energy denominator approximations (CEDA) to the optimized effective potential. The ELP procedure differs from the CEDA and LHF in that it yields the target potential as an expansion in auxiliary basis functions. We report extensive calculations of atomic and molecular properties using the frozen-orbital ELP method and its iterative generalization to prove that ELP results agree with the corresponding LHF and CEDA values, as they should. Finally, we make the case for extending the iterative frozen-orbital ELP method to full orbital relaxation.

The electronic ground-state energy problem: a new reduced density matrix approach.

We present here a formulation of the electronic ground-state energy in terms of the second order reduced density matrix, using a duality argument. It is shown that the computation of the ground-state energy reduces to the search of the projection of some two-electron reduced Hamiltonian on the dual cone of N-representability conditions. Some numerical results validate the approach, both for equilibrium geometries and for the dissociation curve of N(2).