Wednesday, May 4, 2005
12:00 Noon - 1:00 p.m.
3105 Engineering Bldg.
Network topology reveals biological hypotheses
Debra S. Goldberg
Harvard Medical School
Network analysis has become an indispensable tool for understanding complex phenomena as diverse as social relationships, the internet, and chemical reactions in a living cell. Diverse real networks have been shown to share large-scale topological properties. In particular, networks of genes or gene products (proteins) that are systematically determined using high-throughput experimental methods have some well-studied properties. These inaccurate biomolecular networks are increasingly being used to predict protein function, a task that is critical for finding new drug targets and for gaining a systems-level understanding of any organism. To improve the accuracy of protein function prediction, my recent work has exploited network topology to make inferences about genes, proteins, and their relationships.
In this talk, I will show how to predict protein interactions and compute accurate confidence assessments of observed interactions by adapting a measure of small‑world topology. Many networks have hidden subtypes of edges (e.g., true and false positive edges), and I will describe a model selection approach to identify such cases. Finally, I will present several methods that integrate distinct types of data with topological information to predict protein function.