| The rise of genomics, and development of technological advances for the high-throughput study of genome-scale cell activity, is leading to a shift in thinking toward an integrated holistic view of living systems. The elucidation of whole-cell metabolic networks necessitates the development of integrative methods to analyze, interpret, and predict the systemic properties of cellular metabolism. A general constraint-based approach for the comprehensive modeling of metabolic systems is developed and advocated as a paradigm shift in attempting to understand cellular metabolism and the metabolic genotype-phenotype relationship. The consideration of stoichiometric, thermodynamic, and capacity constraints leads to the quantitative assessment of metabolic capabilities and fitness. Utilizing the mathematics of convex analysis, the range of feasible flux distributions, which a network can display, are confined to the steady-state flux cone. This solution space is spanned by a unique set of systemically independent biochemical pathways, termed extreme pathways, based on stoichiometry and limited thermodynamics. These pathways represent the edges of the steady-state flux cone and can be used to represent any flux distribution achievable by the network. These pathways and the constraints-based approach serve to redefine the notion of a metabolic pathway in the context of systemic function, and allow the ability to quantitatively assess metabolic capabilities and fitness. Through the introduction of balanced network demands and capacity constraints, optimal flux distributions can be predicted and interpreted from a pathway-oriented perspective allowing for a deeper understanding of shifts in metabolic resource allocation. These theoretical developments are applied to investigate the metabolic production capabilities and performance of the human red blood cell and Escherichia coli central metabolism, along with Haemphilus influenzae and Helicobacter pylori two bacterial human pathogens whose complete metabolic networks are constructed from genomic, biochemical, and physiological data. From the analysis insight is gained into the structure of the networks, their general fitness, and pathway utilization under changing environmental and genetic conditions. As biology moves into the information age there exists a critical need for integrated theoretical/experimental approaches to study living systems, which is anticipated to place in silico predictive biology as a central component in the advancement of medical and industrial biotechnology. |
