An Onsager-Machlup approach to the most probable transition pathway for a genetic regulatory network

TL;DR

This paper introduces an Onsager-Machlup approach combined with machine learning to identify the most probable transition pathways in gene regulatory networks, analyzing the impact of noise on these transitions.

Contribution

It presents a novel Onsager-Machlup framework integrated with machine learning for transition pathway analysis in gene networks, addressing excitable and bistable dynamics.

Findings

The method effectively identifies transition pathways in gene expression models.

Noise intensity significantly influences transition phenomena.

The approach bridges stochastic analysis and machine learning for biological systems.

Abstract

We investigate a quantitative network of gene expression dynamics describing the competence development in Bacillus subtilis. First, we introduce an Onsager-Machlup approach to quantify the most probable transition pathway for both excitable and bistable dynamics. Then, we apply a machine learning method to calculate the most probable transition pathway via the Euler-Lagrangian equation. Finally, we analyze how the noise intensity affects the transition phenomena.