Dynamic Modeling and Stochastic Simulation of Metabolic Networks

Abstract

Increased technological methods have enabled the investigation of biology at nanoscale levels. Nevertheless, such systems necessitate the use of computational methods to comprehend the complex interactions occurring. Traditionally, dynamics of metabolic systems are described by ordinary differential equations producing a deterministic result which neglects the intrinsic heterogeneity of biological systems. More recently, stochastic modeling approaches have gained popularity with the capacity to provide more realistic outcomes. Yet, solving stochastic algorithms tend to be computationally intensive processes. Employing the queueing theory, an approach commonly used to evaluate telecommunication networks, reduces the computational power required to generate simulated results, while simultaneously reducing expansion of errors inherent to classical deterministic approaches. Herein, we present the application of queueing theory to efficiently simulate stochastic metabolic networks. For the current model, we utilize glycolysis to demonstrate the power of the proposed modeling methods, and we describe simulation and pharmacological inhibition in glycolysis to further exemplify modeling capabilities.\n\nAuthor SummaryComputational biology is increasingly used to understand biological occurances and complex dynamics. Biological modeling, in general, aims to represent a biological system with computational approaches, as realistically and accurate as current methods allow. Metabolomics and metabolic systems have emerged as an important aspect of cellular biology, allowing a more sentive view for understanding the complex interactions occurring intracellularly as a result of normal or perturbed (or diseased) states. To understand metabolic changes, many researchers have commonly used Ordianary Differential Equations to produce in silico models of the in vitro system of interest. While these have been beneficial to date, continuing to advance computational methods of analyzing such systems is of interest. Stochastic models that include randomness have been known to produce more reaslistic results, yet the difficulty and intesive time component urges additional methods and techniques to be developed. In the present research, we propose using queueing networks as a technique to model complex metabolic systems, doing such with a model of glycolysis, a core metabolic pathway.