SEMINAR

March 28 (Fri) 2008, 11AM- SE 213

Predicting protein-protein interactions in highly mutating sequence positions of RNA retroviruses genomes using Mutual Information

Scooter Willis, Department of Computer and Information Science and Engineering, University of Florida, Gainesville, Florida & The Scripps Research Institute

Abstract

All biological functions in the living cell involve the interaction of protein structures in a complex system where through billions of years of evolution the insertion, deletion and mutation of DNA sequences provides the coding for life as we know it. The challenge in understanding this coding is the ability to separate random mutations from positive mutations that are critical to preserving or improving the functional/structural properties of the protein interactions. Mutual Information is used to calculate the mutual dependence between two sequence positions which indicates the information that may describe the importance of compensating mutations.

The application of Mutual Information between two variables is straightforward given that the calculated probability distributions are accurate. Numerous papers have been published on the use of Mutual Information against sequence data to detect compensating mutations with techniques to filter false positives or compensate for the phylogenetic effect associated with using closely related sequence data. The Reduced Phylogenetic Effect (RPE) method is introduced to measure the mutation distributions based on mutations events in the phylogenetic tree resulting in an improved detection of coevolving sequence positions.

The RPE method is used to predict protein-protein interactions as a network topology for HIV, HCV, Influenza A, and Dengue fever which are retroviruses that have high mutation rates, small genomes and (because of the impact of these diseases) are aggressively researched. Current wet lab research techniques used to detect or validate protein interactions in viruses is an evolving field. The use of readily available virus sequence data from wet lab experiments and the RPE method to detect protein interactions can provide invaluable insight into the protein topology or systems biology of a virus.

Contact for more details: rnarayan@fau.edu, URL: http://www.science.fau.edu/fbrcf/