Inferring dissipation from current fluctuations

Todd R. Gingrich, Grant M. Rotskoff, Jordan M. HorowitzJournal of Physics A Mathematical and Theoretical 2017 · DOI · arXiv

Abstract

Complex physical dynamics can often be modeled as a Markov jump process between mesoscopic configurations. When jumps between mesoscopic states are mediated by thermodynamic reservoirs, the time-irreversibility of the jump process is a measure of the physical dissipation. We rederive a recently introduced inequality relating the dissipation rate to current fluctuations in jump processes. We then adapt these results to diffusion processes via a limiting procedure, reaffirming that diffusions saturate the inequality. Finally, we study the impact of spatial coarse-graining in a two-dimensional model with driven diffusion. By observing fluctuations in coarse-grained currents, it is possible to infer a lower bound on the total dissipation rate, including the dissipation associated with hidden dynamics. The tightness of this bound depends on how well the spatial coarse-graining detects dynamical events that are driven by large thermodynamic forces.

Introduction

In the microscopic realm of molecular motors and cellular circuits, systems operate far from thermal equilibrium. A fundamental challenge in biophysics is measuring the "cost" of these processes—the physical dissipation or entropy production. While the classical Fluctuation-Dissipation Theorem (FDT) links fluctuations to response near equilibrium, it fails in the highly driven regimes typical of biology.

Furthermore, we rarely have full visibility into every microscopic state transition. Usually, we only observe "coarse-grained" snapshots of a system's behavior. This paper addresses a critical question: Can we definitively estimate the total energy dissipated by a system if we can only see a small, noisy part of its dynamics?

Theoretical Framework: Markov Jump Processes

The authors model these systems using Markov jump processes, where a system transitions between discrete mesoscopic states. By applying "Level 2.5" large deviation theory, they describe the joint probability of observing specific time-averaged densities and currents over long trajectories.

Steady-state density and current in a Markov jump process on a grid.

The core technical contribution is the derivation of a quadratic bound for the "rate function," which quantifies the likelihood of fluctuations. By projecting this high-dimensional bound onto a single, measurable scalar current, the authors derive the Thermodynamic Uncertainty Relation (TUR). This inequality states that the precision of any measurable current (like particle flow) sets a fundamental lower bound on the system's total entropy production: $$\Sigma^\pi \ge \frac{2(j_d^\pi)^2}{\text{var}(j_d)}$$ where $\Sigma^\pi$ is the total dissipation rate, $j_d^\pi$ is the mean current, and $\text{var}(j_d)$ is its variance.

Validation: The Four-Well Model

To bridge the gap between discrete jumps and continuous physics, the authors use a limiting procedure to apply these bounds to Langevin equations (overdamped diffusion). They validate their theory using a synthetic 2D driven diffusion model.

Potential landscape and external driving force in the 2D four-well diffusion model.

In this model, a particle moves in a potential landscape with four Gaussian wells, driven by a non-gradient external field $f_{\text{ext}}$. By numerically solving the master equation on a fine grid, they compared the predicted dissipation bounds against the known total dissipation.

Key Insights

  1. The Cost of Precision: There is an inescapable trade-off between the precision of a process and its energy cost. High precision (low variance) requires a minimum amount of energy dissipation.
  2. Diffusion Saturates the Bound: For continuous diffusion processes, the inequality becomes an equality, meaning current fluctuations perfectly characterize the dissipation.
  3. Robustness to Coarse-Graining: The bound remains valid even when observing "coarse-grained" signals. For instance, tracking only full rotations of a motor still provides a lower bound on the total entropy produced by all internal chemical steps.
  4. Alignment Matters: The bound is tightest when the measured current is well-aligned with the thermodynamic forces driving the system. In the 2D model, as energy barriers increase (making the system more "jump-like"), the coarse-grained current captures a higher fraction of total dissipation—up to 60% in their tests.

Limitations and Future Outlook

While powerful, the method currently provides only a lower bound. If the chosen observable is poorly aligned with the actual flow, the bound may be uninformative. Additionally, the current derivation assumes steady-state conditions and underlying Markovian dynamics. Future work may extend these principles to transient, time-dependent driving and non-Markovian "hidden" worlds, potentially offering a robust strategy for analyzing real-world single-molecule experimental data.

Summary updated

Abstract

Complex physical dynamics can often be modeled as a Markov jump process between mesoscopic configurations. When jumps between mesoscopic states are mediated by thermodynamic reservoirs, the time-irreversibility of the jump process is a measure of the physical dissipation. We rederive a recently introduced inequality relating the dissipation rate to current fluctuations in jump processes. We then adapt these results to diffusion processes via a limiting procedure, reaffirming that diffusions saturate the inequality. Finally, we study the impact of spatial coarse-graining in a two-dimensional model with driven diffusion. By observing fluctuations in coarse-grained currents, it is possible to infer a lower bound on the total dissipation rate, including the dissipation associated with hidden dynamics. The tightness of this bound depends on how well the spatial coarse-graining detects dynamical events that are driven by large thermodynamic forces.
BibTeX
@article{gingrich2017inferring,
  author = {Todd R. Gingrich and Grant M. Rotskoff and Jordan M. Horowitz},
  title = {{Inferring dissipation from current fluctuations}},
  journal = {Journal of Physics A Mathematical and Theoretical},
  year = {2017},
  doi = {10.1088/1751-8121/aa672f},
}
Figure