Stochastic modeling of excitable dynamics: improved Langevin model for mesoscopic channel noise

TL;DR

This paper presents an improved Langevin model for simulating mesoscopic channel noise in neuronal dynamics, providing better accuracy and computational efficiency compared to previous models.

Contribution

The authors develop a refined Langevin model with natural boundary conditions that accurately captures channel noise variance in Hodgkin-Huxley dynamics.

Findings

The improved Langevin model aligns well with discrete state models.

Previous models by Fox and Lu are effective with boundary conditions.

The new model enhances simulation accuracy for mesoscopic channel noise.

Abstract

Influence of mesoscopic channel noise on excitable dynamics of living cells became a hot subject within the last decade, and the traditional biophysical models of neuronal dynamics such as Hodgkin-Huxley model have been generalized to incorporate such effects. There still exists but a controversy on how to do it in a proper and computationally efficient way. Here we introduce an improved Langevin description of stochastic Hodgkin-Huxley dynamics with natural boundary conditions for gating variables. It consistently describes the channel noise variance in a good agreement with discrete state model. Moreover, we show by comparison with our improved Langevin model that two earlier Langevin models by Fox and Lu also work excellently starting from several hundreds of ion channels upon imposing numerically reflecting boundary conditions for gating variables.