Approximation Techniques for Stochastic Analysis of Biological Systems

TL;DR

This paper introduces a scalable approximation method for probabilistic model checking of biological systems, enabling analysis of complex genetic circuits by reducing state space while maintaining accuracy.

Contribution

It presents a novel state-space approximation technique that improves scalability in probabilistic analysis of biological systems, specifically genetic circuits.

Findings

The method effectively reduces state space size.

It maintains high accuracy compared to existing tools.

Demonstrated on a genetic toggle switch design.

Abstract

There has been an increasing demand for formal methods in the design process of safety-critical synthetic genetic circuits. Probabilistic model checking techniques have demonstrated significant potential in analyzing the intrinsic probabilistic behaviors of complex genetic circuit designs. However, its inability to scale limits its applicability in practice. This chapter addresses the scalability problem by presenting a state-space approximation method to remove unlikely states resulting in a reduced, finite state representation of the infinite-state continuous-time Markov chain that is amenable to probabilistic model checking. The proposed method is evaluated on a design of a genetic toggle switch. Comparisons with another state-of-art tool demonstrates both accuracy and efficiency of the presented method.