Bala Krishnamoorthy, Department of Mathematics, Washington State University
Computational approaches for protein mutagenesis
Abstract: Mutagenesis is the process of changing one or more amino acids (AAs) in a protein to alternative AAs such that the resulting mutant protein has desirable properties the original one lacks, e.g., increased or decreased stability, reactivity, solubility, or temperature sensitivity. Biochemists often have to choose a small subset of mutations from a large set of candidates in order to identify the desirable ones. Computational approaches are invaluable for efficiently making these choices. This talk will overview my work on computational approaches to predict stability and solubility mutagenesis, and to understand the mechanisms of temperature-sensitive (Ts) mutations. I employ techniques from several areas including computational geometry, optimization, machine learning, and statistics. The scoring functions for predicting changes to stability and solubility due to mutations are developed using the computational geometry construct of Delaunay tessellation. In the case of solubility, we optimize the scoring function using linear programming techniques. Our findings for the case of Ts mutations reveal that attributes of the neighborhood of the mutation site are as significant in determining Ts mutants as the properties of the site itself.
