Complex networks arise in fields as disparate as sociology, computer science and biology. In the past decade, extensive research into the properties of the networks of interaction underlying complex systems has uncovered surprising commonalities among the topologies of different systems, hinting at possible shared principles in the way these networks change in time. After reviewing highlights of the general analysis of complex networks (network physics), this talk will focus on network-based dynamic models of biological systems.
A dynamic description of a system's behavior incorporates the network of interactions through the functions that link the state variables of the nodes. How much does the topology of the network of interactions determine the system's dynamics? Can a discrete model with a detailed accounting of the interaction network and a coarse-grained description of the kinetics of individual interactions be successful and predictive? I will present several case studies in diverse biological fields: drought signal transduction in plants, pathogen-immune response interactions in mammals, LGL leukemia in humans. Each model is based on a reconstructed network of interactions and uses the available information on the time-dependent activity or regulation of the nodes. The models reproduce the known sequence of events and predict the key mediators of each process. Several of these predictions were validated experimentally. The success of the models suggests that indeed the topology of the network of interactions plays a determinant role in the system's behavior, and that predictive modeling is possible even in the absence of quantitative kinetic information.
Coffee and cookies before the presentation at 3:15 pm, and refreshments after the presentation will both be served in Room 128.