A Systems Biology Approach to Automatic Pathway Identification

Professor Peter Woolf

University of Michigan

Department of Chemical Engineering

The “-omics” revolution in molecular biology has promised to improve human health by cataloging a vast number of potential gene regulatory pathways involved in disease. However, identification and analysis of these pathways has become the next major challenge, as most diseases are a result of a complex network of interactions. The emerging field of systems biology is well poised to help resolve this issue, as it focuses on developing predictive, systems level models of how biological pathways function.

For this seminar, I will pose this problem as a challenge of system identification. For this I will focus on the following three areas related to systems biology: 1) de novo signal transduction pathway identification using machine learning tools such as Bayesian networks; 2) automated experimental design; and 3) high throughput microfluidics based experimentation. By merging these three technologies, I will discuss the directions my group is taking that we believe is particularly suited to the field of systems biology. As a particular example, I will focus on identifying pathways responsible for embryonic stem cell differentiation and proliferation.

Friday, January 20, 2006

3:30 – 4:30 p.m.

1500 EECS