Scale vs. Detail: Modeling Metabolic Networks Using Varying Levels of Complexity

Computational analysis of biological systems aims to predict cell behavior under different environmental and genetic contexts. Due to the complexity of living organisms, it is not possible to construct a comprehensive mathematical description of context-specific cellular phenotypes. However, with the advent of genome sequencing techniques and the accelerated rate of gene function discovery, the phenotypic landscape of an organism can be determined with the aid of metabolic network reconstructions. These networks describe the biochemical transformations occurring in the cell as a system of mathematical equations. By utilizing concepts from optimization and numerical analysis, different modeling techniques can be used to solve this system of equations. Recently, new frameworks aiming to improve the predictive accuracy of conventional models through incorporation of additional physical and biological considerations have been proposed. In this dissertation, I demonstrate the utility of such frameworks and discuss how they fit within the overall landscape of metabolic modeling. In the first chapter, I highlight the underlying concepts that form the backbone of each framework. In subsequent chapters, I illustrate how these concepts can be utilized to answer biologically relevant questions. The research works presented in this dissertation serve as a stepping-stone towards achieving the ultimate goal of systems biology