Modeling the Sense of Smell with Biological Receptor Arrays
Speaker: Alex Morozov, Rutgers University
Location: Warren Weaver Hall 1302
Date: December 4, 2009, 11:30 a.m.
Host: Dennis Shasha
Receptors involved in intercellular communications are often exquisitely tuned to bind a specific ligand while excluding all others. In contrast, receptors responsible for reading environmental signals must respond to a broader range of stimuli than would be feasible if each stimulus required its own receptor. For example, the mammalian visual system distinguishes a wide variety of colors using only three receptors, while the olfactory system can differentiate between thousands of compounds using several hundred receptors. The discriminatory powers of the visual and olfactory systems are based on a combinatorial mechanism of perception in which receptors have overlapping specificities.
We have sought to model olfactory perception using an artificial receptor array based on G protein-coupled receptors (GPCRs). GPCRs form a class of biochemical sensors with an enormous diversity of ligand binding affinities. We have employed directed evolution in yeast to create novel GPCR-based receptors with broadly overlapping specificities. We demonstrate using Bayesian analysis of biophysical models of receptor-ligand binding that our receptor array can be effectively used to discriminate between constituent compounds in a complex mixture of chemicals. We use information theory to formalize the notion of information capacity of receptor arrays, and to compare the chemical sensing ability of specific and non-specific systems.
Alexandre V. Morozov has a Ph.D. from the Department of Physics, University of Washington, Seattle. Alexandre's postdoctoral research was done at the Center for Studies in Physics and Biology, Rockefeller University, New York with Prof. Eric Siggia. Since September 2007 Alexandre has held a position of an Assistant Professor at the Department of Physics & Astronomy and a faculty member of the BioMaPS Institute for Quantitative Biology, Rutgers University. His current research focuses on computational predictions of chromatin structure and its effect on gene regulation and other cellular functions, on evolutionary interplay between protein fitness, stability and patterns of amino acid mutations, and on mechanisms of protein-DNA and protein-ligand recognition. Dr. Morozov is a past recipient of a Leukemia and Lymphoma Society Postdoctoral Fellowship and a current recipient of an Alfred P. Sloan Research Fellowship. Research in Dr. Morozov's group is supported by the National Institutes of Health.
Refreshments will be offered starting 15 minutes prior to the scheduled start of the talk.