arXiv subjects

Yeongjin Kim

Publications and source records attributed to Yeongjin Kim.

6 recordsLinked to original sources

arXiv2026

Power spectral density of trajectories of active Ornstein-Uhlenbeck particles

The power spectral density (PSD) is a central frequency-domain descriptor of stochastic processes. While PSDs have been studied for Brownian motion and a few anomalous diffusion processes, the spectral densities of active nonequilibrium processes remain almost unexplored. Here, we present an exact theory for the PSDs of active diffusion using the model of active Ornstein-Uhlenbeck particles (AOUPs). We investigate the spectral densities of AOUPs in free space and under harmonic confinement. In free space, active motion does not alter the Brownian $f^{-2}$ spectrum, but only modifies its amplitude and introduces a crossover at the persistence frequency. Under confinement, the spectrum exhibits a rich variety of features depending on the persistence, trap relaxation, and activity strength, including two characteristic signatures that are absent in both thermal systems and free AOUPs. These are a two-plateau structure from a double-trapping mechanism due to two noise sources, and the new $f^{-4}$ spectral scaling associated with transient ballistic motion. We also investigate the finite time effects through the finite-time PSD, and find that the low-frequency plateau and high frequency oscillation exhibit distinct dependences on the observation time $T$ in free and confined systems. Finally, we discuss our results in connection with previously reported experimental studies of active systems. Our results provide an analytically tractable framework for interpreting such systems.

cond-mat.stat-mech↗

arXiv2025

Objective comparison of methods to decode anomalous diffusion

Deviations from Brownian motion leading to anomalous diffusion are found in transport dynamics from quantum physics to life sciences. The characterization of anomalous diffusion from the measurement of an individual trajectory is a challenging task, which traditionally relies on calculating the trajectory mean squared displacement. However, this approach breaks down for cases of practical interest, e.g., short or noisy trajectories, heterogeneous behaviour, or non-ergodic processes. Recently, several new approaches have been proposed, mostly building on the ongoing machine-learning revolution. To perform an objective comparison of methods, we gathered the community and organized an open competition, the Anomalous Diffusion challenge (AnDi). Participating teams applied their algorithms to a commonly-defined dataset including diverse conditions. Although no single method performed best across all scenarios, machine-learning-based approaches achieved superior performance for all tasks. The discussion of the challenge results provides practical advice for users and a benchmark for developers.

physics.data-an↗

arXiv2024

Active Diffusion of Self-Propelled Particles in Semi-Flexible Polymer Networks

Mesh-like structures, such as mucus gel or cytoskeleton networks, are ubiquitous in biological systems. These intricate structures are composed of cross-linked, semi-flexible bio-filaments, crucial to numerous biological processes. In many biological systems, active self-propelled particles like motor proteins or bacteria navigate these intricate polymer networks. In this study, we develop a computational model of three-dimensional cubic-topological, swollen polymer networks of semi-flexible filaments. We perform Langevin dynamics simulations to investigate the diffusion of active tracer particles navigating through these networks. By analyzing various physical observables, we investigate the effects of mesh-to-particle size ratio, Péclet number of active particles, and bending stiffness of the polymer networks upon active trapped-and-hopping diffusion of the tracer. When the tracer size is equal to or larger than the mesh size, the polymer stiffness substantially enhances trapping while suppressing the hopping process. Notably, the mean trapped time exhibits an exponential growth law to the bending stiffness with an activity-dependent slope. An analytic theory based on the mean first-passage time of active particles in a harmonic potential is developed. Our findings deepen the comprehension of the intricate interplay between the polymer's bending stiffness, tracer size, and the activity of tracer particles. This knowledge can shed light on important biological processes, such as motor-driven cargo transport or drug delivery, which hinge on the behavior of active particles within biological gels.

cond-mat.soft↗

arXiv2022

Bayesian inference of scaled versus fractional Brownian motion

We present a Bayesian inference scheme for scaled Brownian motion, and investigate its performance on synthetic data for parameter estimation and model selection in a combined inference with fractional Brownian motion. We include the possibility of measurement noise in both models. We find that for trajectories of a few hundred time points the procedure is able to resolve well the true model and parameters. Using the prior of the synthetic data generation process also for the inference, the approach is optimal based on decision theory. We include a comparison with inference using a prior different from the data generating one.

stat.ME↗

arXiv2022

Active diffusion of self-propelled particles in flexible polymer networks

Biopolymer networks having a meshwork topology, e.g., extracellular matrix and mucus gels, are ubiquitous. It is an open question to understand how self-propelled agents such as Janus colloidal particles diffuse through such a biopolymer network. Here, we computationally explore this issue in-depth by explicitly modeling three-dimensional biopolymer networks and performing Langevin dynamics simulations of active diffusion of the self-propelled tracers therein. We show that the active tracer performs distinct diffusion dynamics depending on the mesh-to-particle size and Péclet number (Pe). When the particle is smaller than the mesh size, it moves as if in free space with a decreased mobility depending on the polymer occupation density and Pe. However, when the particle size is increased to be comparable to the mesh size, the active particles explore the polymer network using the trapped-and-hopping mechanism. We study the trapped time distribution, flight length distribution, the mean-squared displacement, and the long-time diffusivity at varying Pe. If the particle is larger than the mesh, it captures the collective viscoelastic dynamics from the polymer network at short times and the simple diffusion of the total system at large times. Finally, we discuss the scaling behavior of the long-time diffusivity with Pe, where we find a range of Pe that yields a nontrivial power law. The latter turns out to arise from a large fluctuation of trapped, activated tracers in conjugation with responsive polymer networks.

cond-mat.soft↗

arXiv2021

Bayesian inference of Lévy walks via hidden Markov models

The Lévy walk is a non-Brownian random walk model that has been found to describe anomalous dynamic phenomena in diverse fields ranging from biology over quantum physics to ecology. Recurrently occurring problems are to examine whether observed data are successfully quantified by a model classified as Lévy walks or not and extract the best model parameters in accordance with the data. Motivated by such needs, we propose a hidden Markov model for Lévy walks and computationally realize and test the corresponding Bayesian inference method. We introduce a Markovian decomposition scheme to approximate a renewal process governed by a power-law waiting time distribution. Using this, we construct the likelihood function of Lévy walks based on a hidden Markov model and the forward algorithm. With the Lévy walk trajectories simulated at various conditions, we perform the Bayesian inference for parameter estimation and model classification. We show that the power-law exponent of the flight-time distribution can be successfully extracted even at the condition that the mean-squared displacement does not display the expected scaling exponent due to the noise or insufficient trajectory length. It is also demonstrated that the Bayesian method performs remarkably inferring the Lévy walk trajectories from given unclassified trajectory data set if the noise level is moderate.

physics.bio-ph↗