Systems analysis of cellular signaling networks

Cellular signaling networks are at the heart of developmental processes as well as how cells respond to their ever-changing environment. High-throughput technologies are generating data at a genome-scale; thus, large-scale reconstructions of signaling networks are becoming increasingly more feasible and necessary. This doctoral dissertation provides a foundation for the systems level analysis of cellular signaling networks. The first order of magnitude analysis of the human cellular signaling network is presented herein. Furthermore, a novel scheme for the classification of signaling network reconstructions and analyses is discussed. This study is the first to communicate the challenges facing the analysis of signaling networks at a cell-scale with existing approaches. This dissertation presents pioneering efforts that were made in the development of extreme pathway analysis, a tool for the study of biochemical networks. Network-level properties have now been defined with extreme pathway analysis and are presented in this dissertation. For example, pathway redundancy is a quantitative measure of the number of independent routes connecting inputs to outputs. The pathway redundancy in the metabolic network of H. influenzae was evaluated. Correlated reactions sets are groups of reactions that function together and can be calculated from the stoichiometric matrix of a network. They are thus unbiased network modules. The correlated reaction sets in metabolic networks of H. pylori and H. influenzae as well as the JAK-STAT signaling network in the human B cell are presented. Furthermore, the first comparison of extreme pathways and elementary modes (a similar analytical technique) for actual biochemical networks is presented herein. For even small systems these two similar approaches have tremendously complicated relationships. The innovative application of extreme pathway analysis to a prototypic signaling network is also discussed. This study presented novel definitions of crosstalk and input/output relationships in signaling systems. The JAK-STAT signaling network discussed in this dissertation is the largest stoichiometric reconstruction to date. The redundancy in this network was calculated and defined. Furthermore, network crosstalk was evaluated. These analyses are the first ever for signaling networks. This dissertation thus discusses pioneering work in the application of large-scale analysis methods to cellular signaling and metabolic networks.