-----------------------------------------------------------------------
                                           BIOINFORMATICS COLLOQUIUM
                                         School of Computational Sciences
                                              George Mason University
                       -----------------------------------------------------------------------
 
       Computational Mutagenesis for Predicting Functional Consequences of Amino Acid Replacements in Proteins
					        Majid Masso, Ph.D.
       
					     George Mason University

Abstract:
Experimental protein mutagenesis studies that attempt to measure relative changes in 
stability or activity are critical to biochemists for investigating factors involved 
in protein folding, determining functional roles of amino acids, and undertaking targeted 
protein engineering projects. Medical researchers and clinicians are equally interested 
in assaying variant human proteins for their potential association with disease as well as 
mutant target proteins (human, bacterial, or viral) for their potential resistance to inhibitor 
medications. The ability to accurately predict how amino acid replacements impact these 
protein functions would represent a considerable savings of both time and money. In this 
talk, I will detail a computational mutagenesis methodology arising from a four-body, 
knowledge-based statistical contact potential. Application of this mutagenesis has been 
utilized for quantifying environmental perturbations that an amino acid substitution 
imparts on all residue positions in a given protein structure. The data collected for 
the variant proteins form the basis of mutant feature vectors used to train predictive 
models with supervised classification and regression machine learning algorithms. Many 
unexplored research topics will be identified for M.S. or Ph.D. students interested in 
pursuing thesis or dissertation projects in this area.