QSB: Discrete Optimization Techniques for Probing the Performance Limits of Metabolic Networks

The aim of the proposed work is to extend the use of linear and other network analysis tools. The addition of logical information in model constraints to emulate regulatory logic will be pursued. Mixed integer linear programming will be among the tools employed. Two general approaches will be used. The first entails encoding a check to see if a flux solution is consistent with the kinetic and regulatory structure of the network. The second involves superimposing transcript ratio data from DNA arrays within probabilistic framework that account for the lack of linearity between transcript level and gene product activity.