Learning force fields from stochastic trajectories

Abstract

When monitoring the dynamics of stochastic systems, such as interacting particles agitated by thermal noise, disentangling deterministic forces from Brownian motion is challenging. Indeed, we show that there is an information-theoretic bound, the capacity of the system when viewed as a communication channel, that limits the rate at which information about the force field can be extracted from a Brownian trajectory. This capacity provides an upper bound to the system's entropy production rate, and quantifies the rate at which the trajectory becomes distinguishable from pure Brownian motion. We propose a practical and principled method, Stochastic Force Inference, that uses this information to approximate force fields and spatially variable diffusion coefficients. It is data efficient, including in high dimensions, robust to experimental noise, and provides a self-consistent estimate of the inference error. In addition to forces, this technique readily permits the evaluation of out-of-equilibrium currents and the corresponding entropy production with a limited amount of data.

arXiv comment10 pages text + 24 pages appendix

References in corpus (29)

• Double-slit photoelectron interference in strong-field ionization of the neon dimer
• Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
• Stochastic thermodynamics, fluctuation theorems, and molecular machines
• Thermodynamic uncertainty relation for biomolecular processes
• Emergence of macroscopic directed motion in populations of motile colloids
• Dimensionality and dynamics in the behavior of C. elegans
• State-dependent diffusion: thermodynamic consistency and its path integral formulation
• Broken detailed balance and non-equilibrium dynamics in living systems
• Introduction to dynamical large deviations of Markov processes
• Influence of non-conservative optical forces on the dynamics of optically trapped colloidal spheres: The fountain of probability
• Fluctuation Relations for Diffusion Processes
• Steady state statistics of driven diffusions
• Penalized nonparametric mean square estimation of the coefficients of diffusion processes
• On the proper reconstruction of complex dynamical systems spoilt by strong measurement noise
• Many-Body Electrostatic Forces Between Colloidal Particles at Vanishing Ionic Strength
• Nonparametric estimation of scalar diffusions based on low frequency data
• Broken Detailed Balance of Filament Dynamics in Active Networks
• Entropy Production Rate is Maximized in Non-Contractile Actomyosin
• High-Performance Reconstruction of Microscopic Force Fields from Brownian Trajectories
• Eulerian and Lagrangian pictures of non-equilibrium diffusions
• Non-equilibrium scaling behaviour in driven soft biological assemblies
• Noninvasive Measurement of Dissipation in Colloidal Systems
• Real-time calibration of a feedback trap
• Experimental metrics for detection of detailed balance violation
• Inflow rate, a time-symmetric observable obeying fluctuation relations
• Statistics of the stochastically-forced Lorenz attractor by the Fokker-Planck equation and cumulant expansions
• Variational estimation of the drift for stochastic differential equations from the empirical density
• Exact results for a simple epidemic model on a directed network: Explorations of a system in a non-equilibrium steady state
• Manifest and Subtle Cyclic Behavior in Nonequilibrium Steady States

Cited by in corpus (43)

• Estimating entropy production by machine learning of short-time fluctuating currents
• Inferring the dynamics of underdamped stochastic systems
• Learning physical properties of anomalous random walks using graph neural networks
• Soft Metamaterials: Adaptation and Intelligence
• Non-Gaussian diffusion near surfaces
• Building general Langevin models from discrete data sets
• Mathematical, Thermodynamical, and Experimental Necessity for Coarse Graining Empirical Densities and Currents in Continuous Space
• Inference of Markov models from trajectories via Large Deviations at Level 2.5 with applications to random walks in disordered media
• Tuning transduction from hidden observables to optimize information harvesting
• Spectral fingerprints of non-equilibrium dynamics: The case of a Brownian gyrator
• Stochastic Inference of Surface-Induced Effects using Brownian Motion
• Broken detailed balance and entropy production in directed networks
• Thermodynamic Uncertainty Relation in Interacting Many-Body Systems
• Optimal reaction coordinates and kinetic rates from the projected dynamics of transition paths
• Discovering dynamic laws from observations: the case of self-propelled, interacting colloids
• Fluctuating landscapes and heavy tails in animal behavior
• Maximum likelihood estimates of diffusion coefficients from single-particle tracking experiments
• Information-driven transitions in projections of underdamped dynamics
• Bézier interpolation improves the inference of dynamical models from data
• Inferring entropy production in anharmonic Brownian gyrators
• Learning minimal representations of stochastic processes with variational autoencoders
• Irreversibility in linear systems with colored noise
• Large deviations of the stochastic area for linear diffusions
• Variance sum rule: proofs and solvable models
• Estimate of entropy generation rate can spatiotemporally resolve the active nature of cell flickering
• Data-driven physics-based modeling of pedestrian dynamics
• Learning Dynamics from Multicellular Graphs with Deep Neural Networks
• Single-Trajectory Characterization of Active Swimmers in a Flow
• Optimal Work Extraction and the Minimum Description Length Principle
• Stochastic line integrals and stream functions as metrics of irreversibility and heat transfer
• Inverse Design of Non-Equilibrium Steady-States: A Large Deviation Approach
• Jensen bound for the entropy production rate in stochastic thermodynamics
• Physics-informed graph neural networks enhance scalability of variational nonequilibrium optimal control
• Improved estimation for energy dissipation in biochemical oscillations
• Attaining entropy production and dissipation maps from Brownian movies via neural networks
• Exact potentials in multivariate Langevin equations
• Inferring nonequilibrium thermodynamics from tilted equilibrium using information-geometric Legendre transform
• Reconstructing Noisy Gene Regulation Dynamics Using Extrinsic-Noise-Driven Neural Stochastic Differential Equations
• Inferring nonlinear fractional diffusion processes from single trajectories
• Force-free kinetic inference of entropy production
• Coarse-graining nonequilibrium diffusions with Markov chains
• Characterizing nonlinear dynamics by contrastive cartography
• The Limits of Inference in Complex Systems: When Stochastic Models Become Indistinguishable