The transport phenomena of Lennard-Jones molecules through a structural bottleneck driven by an external force are investigated by molecular dynamics simulations. We observe two distinct molecular flow regimes distinguished by a critical external force $F_{c}$ and find scaling behaviors between external forces and flow rates. Below the threshold $F_{c}$, molecules are essentially stuck in the bottleneck due to the attractive interaction between the molecules, while above $F_{c}$, molecules can smoothly move in the pipe. A critical flow rate $q_{c}$ corresponding to $F_{c}$ satisfies a simple relationship with angles and the value of $q_{c}$ can be estimated by a simple argument. We further clarify the role of the temperature dependence in the molecular flows through the bottleneck.
Graph Neural Network-State Predictive Information Bottleneck (GNN-SPIB) approach for learning molecular thermodynamics and kinetics
University of Maryland（马里兰大学）；Institute for Physical Science and Technology（物理科学与技术学院）；University of Maryland Institute for Health Computing（马里兰大学健康计算学院）
Combined Representation and Generation with Diffusive State Predictive Information Bottleneck
Department of Physics（物理系）；Institute for Physical Science and Technology（物理科学与技术研究院）；Institute for Health Computing（健康计算研究院）；Biophysics Program（生物物理项目）；Department of Chemistry and Biochemistry（化学与生物化学系）；University of Maryland（马里兰大学）
The diffusion of a molecular motor in the presence of a constant external force is considered on the basis of a simple theoretical model. The motor is represented by a Brownian particle moving in a series of parabolic potentials placed periodically on a line, and the potential is switched stochastically from one parabola to another by a chemical reaction, which corresponds to the ATP hydrolysis or synthesis in motor proteins. It is found that the diffusion coefficient as a function of the force exhibits peaks. The mechanism of this diffusion enhancement and the possibility of observing it in F…
The ability to make sense of the massive amounts of high-dimensional data generated from molecular dynamics (MD) simulations is heavily dependent on the knowledge of a low dimensional manifold (parameterized by a reaction coordinate or RC) that typically distinguishes between relevant metastable states and which captures the relevant slow dynamics of interest. Methods based on machine learning and artificial intelligence have been proposed over the years to deal with learning such low-dimensional manifolds, but they are often criticized for a disconnect from more traditional and physically int…
It has recently been pointed out that Hamiltonian particle systems in constant magnetic fields satisfy generalized time-reversal symmetries that enable to prove useful statistical relationships based on equilibrium phase-space probability distributions without the need to invert, as commonly considered necessary, the magnetic field. Among these relations, that hold without need of Casimir modifications, one finds the standard linear response Green-Kubo relations, and consequently the Onsager reciprocal relations. Going beyond linear response is also possible, for instance in terms of transient…
We investigate the motion of a colloidal particle driven out of equilibrium by an external torque. We use the molecular dynamics simulation that is alternative to the numerical integration approach based on the Langevin equation and is expected to mimic an experiment more realistically. We choose a heat bath composed of thousands of particles interacting to each other through the Lennard-Jones potential and impose the Langevin thermostat to maintain it in equilibrium. We prepare a single colloidal particle to interact with the particles of the heat bath also by the Lennard-Jones potential whil…
We propose an open-boundary molecular dynamics method in which an atomistic system is in contact with an infinite particle reservoir at constant temperature, volume and chemical potential. In practice, following the Hamiltonian adaptive resolution strategy, the system is partitioned into a domain of interest and a reservoir of non-interacting, ideal gas, particles. An external potential, applied only in the interfacial region, balances the excess chemical potential of the system. To ensure that the size of the reservoir is infinite, we introduce a particle insertion/deletion algorithm to contr…
