Molecular Dynamics Simulation Algorithm Research Articles

Sampling rare conformational transitions with a quantum computer

Structural rearrangements play a central role in the organization and function of complex biomolecular systems. In principle, Molecular Dynamics (MD) simulations enable us to investigate these thermally activated processes with an atomic level of resolution. In practice, an exponentially large fraction of computational resources must be invested to simulate thermal fluctuations in metastable states. Path sampling methods focus the computational power on sampling the rare transitions between states. One of their outstanding limitations is to efficiently generate paths that visit significantly different regions of the conformational space. To overcome this issue, we introduce a new algorithm for MD simulations that integrates machine learning and quantum computing. First, using functional integral methods, we derive a rigorous low-resolution spatially coarse-grained representation of the system’s dynamics, based on a small set of molecular configurations explored with machine learning. Then, we use a quantum annealer to sample the transition paths of this low-resolution theory. We provide a proof-of-concept application by simulating a benchmark conformational transition with all-atom resolution on the D-Wave quantum computer. By exploiting the unique features of quantum annealing, we generate uncorrelated trajectories at every iteration, thus addressing one of the challenges of path sampling. Once larger quantum machines will be available, the interplay between quantum and classical resources may emerge as a new paradigm of high-performance scientific computing. In this work, we provide a platform to implement this integrated scheme in the field of molecular simulations.

GPU-accelerated molecular dynamics simulation of solid covalent crystals

Graphics processing unit (GPU) is becoming a powerful computational tool in science and engineering. In this paper, different from previous molecular dynamics (MD) simulation with pair potentials and many-body potentials, two MD simulation algorithms implemented on a single GPU are presented to describe a special category of many-body potentials – bond order potentials used frequently in solid covalent materials, such as the Tersoff potentials for silicon crystals. The simulation results reveal that the performance of GPU implementations is apparently superior to their CPU counterpart. Furthermore, the proposed algorithms are generalised, transferable and scalable, and can be extended to the simulations with general many-body interactions such as Stillinger–Weber potential and so on.

Time dependence of phase variables in a steady shear flow algorithm

We study the periodic time dependence of shear stress that occurs in a low- and a high-density fluid in a molecular dynamics algorithm for simulation of constant shear flow. We present a generalization of the linear response theory for a case when the equilibrium relaxation function and the equilibrium shear stress depend on strain. The predictions of this generalization reveal that the time dependence at low densities is a completely nonlinear effect. At high densities the amplitude of equilibrium oscillations of shear stress with strain is modified by strain-dependent viscosity, causing a decrease in amplitude proportional to shear rate in the linear response regime.

Effects of dispersion forces on the structure and thermodynamics of fluid krypton

Semianalytical and numerical calculations are performed to predict the structural and thermodynamic properties of low-density Kr fluid. Assuming that the interatomic forces can be modelled by a pairwise potential plus the three-body Axilrod-Teller potential, two different routes are explored. The first one is based on the hybridized mean spherical approximation integral equation of the theory of liquids and the second one uses large-scale molecular dynamics (MD). Algorithms for MD simulation are constructed on parallel machines to reduce the amount of computer time induced by the calculations of the three-body forces and the pair-correlation function. Our results obtained with the two methods mentioned above are in quite good agreement with the recent small-angle neutron-scattering experiments [Formisano et al., Phys. Rev. Lett. 79, 221 (1997); Benmore et al., J. Phys.: Condens. Matter 11, 3091 (1999)]. Moreover, the reliability of the asymptotic form of the integral equation is assessed for the specific case of dispersion forces including the three-body contributions, by an analysis at low wave vector and low density. It is seen that the effects of the Axilrod-Teller triple-dipole potential cannot be ignored to describe the structure and the thermodynamic properties of fluid krypton even at low density.

A molecular dynamics algorithm for simulation of field theories in the canonical ensemble

We add a single scalar degree of freedom (“demon”) to the microcanonical ensemble which converts its molecular dynamics into a simulation method for the canonical ensemble (euclidean path integral) of the underlying field theory. This generalization of the microcanonical molecular dynamics algorithm simulates the field theory at fixed coupling with a completely deterministic procedure. We discuss the finite size effects of the method, the equipartition theorem and ergodicity. The method is applied to the planar model in two dimensions and SU(3) lattice gauge theory with four species of light, dynamical quarks in four dimensions. The method is much less sensitive to its discrete time step than conventional Langevin equation simulations of the canonical ensemble. The method is a straightforward generalization of a procedure introduced by S. Nośe for molecular physics.