Data Integration with Constraint-Based Genome-Scale Models

Genome-scale modeling is a powerful tool for quantifying relationships between genetic, metabolic, and phenotypic factors. The mechanistic detail of these models presents unique opportunities in metabolic engineering, drug discovery, and disease comprehension. Despite recent advances, several challenges remain for the genome-scale modeling field. Existing models focus almost exclusively on metabolism, and few studies integrate other biochemical systems. Models describe all functionality encoded in an organism’s genome, and tailoring models to specific genetic states is a difficult, highly specialized process. Large-scale models require equally large datasets for validation, yet few comprehensive datasets exist (especially for lesser-studied organisms).

This thesis focuses on developing methods to overcome the limitations of current genome-scale modeling techniques. Herein, I describe novel methods to

– assemble, visualize, and simulate large models
– design instrumentation to rapidly produce genome-scale datasets for model validation
– improve algorithms to leverage high-throughput expression data and refine models for a particular condition

Together, these advances provide a framework for contextualizing high-throughput data with genome-scale modeling. Special attention is given to the interpretation of functional genomics data through the enzymatic architecture underlying the metabolic network.