Novel Algorithms and Tools for Computational Protein Design with Applications to Drug Resistance Prediction, Antibody Design, Peptide Inhibitor Design, and Protein Stability Prediction

Abstract

Proteins are biological macromolecules made up of amino acids. Proteins range from enzymes to antibodies and perform their functions through a variety of mechanisms, including through protein-protein interactions (PPIs). Computational structure-based protein design (CSPD) seeks to design proteins toward some specific or novel function by changing the amino acid composition of a protein and modeling the effects. CSPD is a particularly challenging problem since the size of the search space grows exponentially with the number of amino acid positions included in each design. This challenge is most often encountered when considering large designs such as the re-design of a PPI. Herein, we discuss how to use CSPD to predict resistance mutations in the active site of the dihydrofolate reductase enzyme from methicillin-resistant Staphylococcus aureus and we investigate the accuracy of an existing CSPD suite of algorithms, osprey. We have also developed novel algorithms and tools within osprey to more efficiently and accurately predict the effects of mutations. We apply these various algorithms and tools to three systems toward a variety of goals: predicting the affect on stability of mutations in staphylococcal protein A (SpA), re-designing HIV-1 broadly neutralizing antibody PG9-RSH toward improved potency, and designing toward a peptide inhibitor of KRas:effector PPIs.