New Computational Methods to Predict Cancer Resistance Mutations and Design D-Peptide Therapeutics

Abstract

Improving disease treatment relies on advancements in our understanding of disease etiology and evolution. Rational drug design seeks to exploit this understanding to improve human health through targeted molecular interventions. In this dissertation, we present computational methods that 1) predict disease evolution in the form of resistance mutations; and 2) generate de novo D-peptide therapeutics. First, we introduce the Resistor algorithm. Resistor uses Pareto optimization with multistate design and cancer-specific mutational probabilities to rank resistance mutations based on their ability to ablate binding to an inhibitor, retain native function, and occur in a specific cancer type. We apply Resistor to 8 inhibitors targeting the EGFR, BRAF, and ERK2 proteins, and provide experimental validation of Resistor-predicted resistance mutations. Second, we introduce DexDesign, a novel algorithm for computationally designing de novo D-peptide inhibitors. DexDesign leverages three novel techniques that are broadly applicable to computational protein design: the Minimum Flexible Set, K*-based Mutational Scan, and Inverse Alanine Scan. We apply these techniques and DexDesign to generate novel D-peptide inhibitors of two biomedically important PDZ domain targets: CALP and MAST2. Notably, the peptides we generated are predicted to bind their targets tighter than their targets' endogenous ligands, validating the peptides' potential as lead therapeutic candidates. We provide implementations of Resistor and DexDesign in the free and open source computational protein design software OSPREY.