Experiment design for systems biology

Mechanism-based chemical kinetic models are increasingly being used to describe biological signaling. Such models serve to encapsulate current understanding of pathways and to enable insight into complex biological processes. Despite the growing interest in these models, a number of challenges frustrate the construction of high-quality models. First, the chemical reactions that control biochemical processes are only partially known, and multiple, mechanistically distinct models often fit all of the available data and known chemistry. We address this by providing methods for designing dynamic stimuli that can distinguish among models with different reaction mechanisms in stimulus-response experiments. We evaluated our method on models of antibody-ligand binding, mitogen-activated protein kinase phosphorylation and de-phosphorylation, and larger models of the epidermal growth factor receptor (EGFR) pathway. Inspired by these computational results, we tested the idea that pulses of EGF could help elucidate the relative contribution of different feedback loops within the EGFR network. These experimental results suggest that models from the literature do not accurately represent the relative strength of the various feedback loops in this pathway. In particular, we observed that the endocytosis and feedback loop was less strong than predicted by models, and that other feedback mechanisms were likely necessary to deactivate ERK after EGF stimulation. Second, chemical kinetic models contain many unknown parameters, at least some of which must be estimated by fitting to time-course data. We examined this question in the context of a pathway model of EGF and neuronal growth factor (NGF) signaling. Computationally, we generated a palette of experimental perturbation data that included different doses of EGF and NGF as well as single and multiple gene knockdowns and overexpressions. While no single experiment could accurately estimate all of the parameters, we identified a set of five complementary experiments that could. These results suggest that there is reason to be optimistic about the prospects for parameter estimation in even large models. Third, there is no standard formulation for chemical kinetic models of biological signaling. We propose a general and concise formulation of mass action kinetics based on sparse matrices and Kronecker products. This formulation allows any mass action model and its partial derivatives to be represented by simple matrix equations, which enabled straightforward application of several numerical methods. We show that models that use other rate laws such as MichaelisMenten can be converted to our formulation. We demonstrate this by converting a model of Escherichia coli central carbon metabolism to use only mass action kinetics. The dynamics of the new model are similar to the original model. However, we argue that because our model is based on fewer approximations it has the potential to be more accurate over a wider range of conditions. Taken together, the work presented here demonstrates that experimental design methodology can be successfully used to improve the quality of mechanism-based chemical kinetic models.

Department dc:contributor.department

Massachusetts Institute of Technology. Dept. of Biological Engineering.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2009

Author dc:creator

• Apgar, Joshua Farley

Advisor dc:contributor.advisor

• Bruce Tidor and Forest M. White.

• Biological Engineering.

Statement dc:rights

• MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/61217

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/61217