Rudiyanto Gunawan
Dep. of Chemistry and Applied Bioscience
ETH Zürich
Switzerland
The maintenance of normal cellular functions and the regulation of cellular responses to environmental and genetic perturbations depend critically on a network of interactions among biomolecules. The reconstruction of biological interaction networks from molecular biology data has become a major focus in systems biology. Despite intense efforts, the inference of network structure and parameters from such data is still unsolved. This inference problem has often been said to be underdetermined, suggesting that there exist indistinguishable solutions. In this seminar, I will describe methods that my group has recently developed to address the underdetermined nature of the biological network inference problem. More specifically, we have adopted an ensemble modeling strategy in which we infer a family of network models that are indistinguishable based on the available data. The first method, called Transitive Reduction and Closure Estimation (TRaCE), has been developed for the inference of an ensemble of gene regulatory network (GRN) structures from differential gene expression data. We have used TRaCE to analyze the inferability of GRN from single- and double-gene knock-out experiments. The second method has been developed for the inference of kinetic parameters from time-series data, which we have used for the creation of kinetic differential equation models of metabolic networks. Here, we formulated the parameter inference as a nested optimization to reduce the problem complexity. In this case, we can define the model ensemble in a lower dimensional parameter space. The second method is available as a MATLAB toolbox called Reduced Dimension Ensemble Modeling and Parameter Estimation (REDEMPTION). Finally, I will present our current efforts in using an ensemble of network models to design (optimal) experiments using a Bayesian approach.
Rudiyanto Gunawan is an Assistant Professor of Chemical and Biological Systems Engineering in the Institute for Chemical and Bioengineering at ETH Zurich, Switzerland. His research interests primarily lie at the intersection of systems engineering and biology, more specifically systems modeling and analysis of cellular networks. His research group activities focus on the development of methods for model identification and analysis of gene regulatory networks, signal transduction pathways, and metabolic networks. In the past six years, Prof. Gunawan has also cultivated a deep interest in the area of biogerontology, specifically, the formation and accumulation of mitochondrial DNA mutations. His research work has resulted in more than 50 peer-reviewed journal and conference papers, 3 book chapters and 1 patent. Prof. Gunawan is a co-recipient of two Best Paper awards from the Journal of Process Control and Computers and Chemical Engineering journal in 2008.