Hearing the Shape of Cancer: Spectral Graph Theory for Pathway Analysis of Gene Expression Data


Wednesdays@NICO Seminar, Noon, October 24th, Chambers Hall, Lower Level

Prof. Rosemary Braun, Northwestern University


High-throughput gene expression profiling assays provide detailed measurements of the expression of thousands of genes per sample. To better understand complex diseases, we wish to make inferences about biological regulation dynamics by considering expression data in the context of interaction networks.  This talk presents a novel approach to pathway analysis.  Using edge weights derived from experimental measurements and putative network topologies from curated pathway databases, we compute the spectral characteristics of the resulting weighted graphs.  I will demonstrate how analysis of the pathway spectra can be used to identify significantly altered systems without relying on single-gene associations, and more importantly, how it can enable inferences about a pathway's dynamical properties based on static gene expression measurements.


Dr. Rosemary Braun is Assistant Professor of Biostatistics in the Department of Preventive Medicine and Lurie Cancer Center at Northwestern University.  Her research focuses on computational and mathematical modeling of biological systems, with the goal of elucidating interactions underlying complex diseases.  She obtained a Physics PhD in 2004 from the University of Illinois at Urbana-Champaign, where she worked with Klaus Schulten on computational biophysics methods.  Following her PhD, she pursued an MPH at Johns Hopkins, where her work with Giovanni Parmigiani shifted her interests from molecular dynamics to high-dimensional data analysis.  Dr. Braun continued her computational biostatistics research as a Cancer Prevention Fellow at the National Cancer Institute.  She joined Northwestern in October 2011.