UC San Diego

Constraint Based Reconstruction and Analysis (COBRA) is a framework within the field of Systems Biology which aims to understand cellular metabolism through the analysis of large scale metabolic models. These models are based on meticulously curated reconstructions of all chemical reactions in an organism. Instead of attempting to predict the exact state of the biological system, COBRA describes the physiological constraints that the system must satisfy and studies the range of solutions satisfying these constraints. Monte Carlo Sampling is one of the COBRA methods used to study how biological properties are distributed over the entire solution space. A set of randomly distributed solutions is generated and serves as a proxy for the entire space. Various aspects of Monte Carlo Sampling in Systems Biology are illustrated : 1) Monte Carlo Sampling has been used historically (Chapter 1), 2) A faster and more efficient procedure for generating Monte Carlo Samples is developed (Chapter 2); 3) Carbon 13 tracing experiments are an important tool for measuring reaction rates through a network. Monte Carlo Sampling was used to optimize the choice of label and explain and measure the dimensionality of output data (Chapter 3); 4) It is possible to incorporate the thermodynamic "loop-law'' into many COBRA methods including sampling (Chapter 4). Additionally two software projects are presented which assist in analyzing COBRA models : 1) the BiGG knowledgebase of reconstructions (Chapter 5) and 2) the COBRA Matlab toolbox v. 2.0 (Chapter 6). These two projects make COBRA methods available to the scientific community