Biophysical fitness landscape design traps viral evolution

We introduce foundational principles for designing customizable fitness landscapes for proteins. We focus on crafting antibody ensembles to create evolutionary traps which restrict viral fitness enhancement. By deriving a fundamental relationship between a mutant protein’s fitness and its binding affinities to host receptors and antibodies, we show that the fitnesses of different protein sequences are designable, meaning they can be independently tuned by careful choice of antibodies. Given a user-defined target fitness landscape, stochastic optimization can be performed to obtain such an ensemble of antibodies which force the protein to evolve according to the designed target fitness landscape. We conduct in silico serial dilution experiments using microscopic chemical reaction dynamics to simulate viral evolution and validate the fitness landscape design. We then apply the design protocol to control the relative fitnesses of two SARS-CoV-2 neutral genotype networks while ensuring absolute fitness reduction. Finally, we introduce an iterative design protocol which consistently discovers better vaccination target sequences, generating antibodies that restrict the post-vaccination fitness growth of escape variants while simultaneously suppressing wildtype fitness. Biophysical fitness landscape design thus opens the door to prescient vaccine, antibody, and peptide design, thinking several steps ahead of pathogen evolution.