Event
Advanced Networks Colloquium: Michelle Girvan, "Network Structure of Annotation Data"
Friday, December 2, 2011
11:00 a.m.
1146 A.V. Williams Building
Kimberly Edwards
301 405 6579
kedwards@umd.edu
Advanced Networks Colloquium
Using the Network Structure of Annotation Data to Gain Insights into Gene Interactions and the Organization of Biological Function
Michelle Girvan
Assistant Professor
Department of Physics and Institute for Physical Science and Technology
University of Maryland
| video |
Abstract
The Gene Ontology (GO) provides a controlled vocabulary of terms for describing gene functions and specifies how these functional terms are related to each other. Biologists can then submit annotations connecting genes to the appropriate functional terms. We propose a method for using the network structure of gene-term annotations associated with GO to investigate the implications of functional similarity for gene interactions. We also exploit the annotation network to explore the organization of biological function.
If one gene regulates another, those two genes are likely to be involved in many of the same biological functions. Conversely, shared biological function may be suggestive of the existence and nature of a regulatory interaction. With this in mind, we develop a new measure of functional similarity between genes based on annotations made to the Gene Ontology (GO) in which the magnitude of their functional relationship is also indicative of a regulatory relationship. We observe that the strength of our similarity measure is correlated with the structural importance of links in the known regulatory network. Our results suggest that structural features in the annotation network can be used to infer and classify regulatory interactions between genes.
To gain insights into gene interactions, we study gene-gene relationships in the annotation network. By contrast, in order to investigate the organization of biological function, we use term-term relationships in the annotation network to establish an alternate natural grouping of biological functions which is significantly different from the conceptual hierarchical structure that relates functional terms in the ontology. Grouping terms by our alternate scheme provides a new framework with which to describe and predict the functions of experimentally identified groups of genes. We discuss the differences between the functional groupings we identify from the annotation network and the traditional division of functions.
Biography
Michelle Girvan is an assistant professor in the Department of Physics and the Institute for Physical Science and Technology at the University of Maryland. Her research interests focus on the theory and applications of complex networks. In terms of theory, she has worked extensively on defining and identifying modularity in networks. She currently works on applications of network theory to gene regulation and the spread of ideas. Girvan received her Ph.D. in physics from Cornell University and received undergraduate degrees in physics and math at MIT. Before joining the faculty at the University of Maryland, she was a postdoctoral fellow at the Santa Fe Institute.
