New methods for modeling large-scale biochemical networks

In this thesis, we explore new modeling approaches that can be used to analyze the systems properties of the biochemical networks of living organisms. Each of the approaches presented has particular advantages over the classical, deterministic, mass-action kinetics models. Stochastic simulations, for example, capture the effects of intrinsic fluctuations when small numbers of proteins are present in a network. Although analytic solutions are not generally available for most models, it is possible to obtain them for open, boundary-driven, linear systems via the grand canonical model. Most biochemical systems, however, involve complex pathways and, in most cases, it is not possible to gather detailed kinetic rate information for each pathway. In such cases, stoichiometric constraints-based optimization approaches provide a means of quantitatively analyzing the possible phenotypes of complex networks without requiring detailed kinetic rate information. Also of interest are the effects of environmental and internal changes on a system, which can be studied by using the methods of sensitivity and control analysis. Each of these methods aims to enhance our understanding of complex biochemical networks and brings us closer to developing a complete model of a living organism.