| The inability to adequately model and simulate many complex dynamic systems results from a lack of theoretical and computational tools that are capable of handling the increased complexity of these systems. This complexity contributes to our inability to identify appropriate mathematical representations that would advance the conceptual and theoretical understanding of these systems. This is particularly evident when attempting to understand the emergence of complex biological phenomenon from interactions that span many levels of biological organization. The complex properties of these systems limit the depth of understanding that can be acquired by more traditional trial and error, reductionists, and experimental analyses. Systems biology focuses on the development of specialized theoretical, analytical, and computational tools that can facilitate a paradigm shift in how we understand, model, manipulate and control biological processes.;The research outlined in this dissertation focuses on the development, refinement, and implementation of engineering methodologies and design concepts for systems biology applications. The work outlined here focuses on two approaches. The first was a procedure formulated around blind source separation techniques. This approach exploited assumptions about underlying biological characteristics in order to blindly extract hidden functional relationships embedded in large-scale gene expression data. The second approach presented an estimation algorithm that used an optimal control formulation to address biological constraints on the component and pathway dynamics. This type of formulation allowed for the direct integration of dynamic equality and inequality constraints into the solution, alleviating the need for constrained optimization routines.;The approaches presented here take advantage of fundamental analytical concepts in electrical engineering. Application specific modifications facilitate the integration of these analytical concepts with available experimental data and first-principles knowledge. This adds to the library of available tools that are capable of estimating and understanding the unknown or immeasurable characteristics of biological processes. Ultimately, these tools will help researchers attain the medicinal, nutritional, and social benefits that can be achieved through the systematic manipulation of these processes at various levels of biological organization. |
