Young Investigator Award: Statistical Learning of Viral Fitness Landscapes for in silico Vaccine Design

The coupling of â??big computingâ? â?? petascale systems and the multicore paradigm â?? with â??big biologyâ? â?? high-throughput sequencing â?? present new opportunities for statistical, molecular, and data driven modeling of biological systems. We have developed an approach rooted in spin glass models from statistical physics to translate clinical sequence databases of hepatitis C virus into quantitative molecular models of viral fitness. Mapping out the replicative fitness of the virus as a function of its amino acid sequence, these empirical fitness landscapes prescribe the mutational playing field over which the virus is constrained to evolve. Exploiting the isomorphism of these landscapes to free energy surfaces, we have identified particular combinations of mutations that can induce a phase transition and collapse of the viral population. Agent-based predator-prey simulations of the viral dynamics over the fitness landscape under host immune pressure can identify immune responses capable of crippling viral fitness. Together these approaches reveal vulnerable immune targets in the viral proteome for the rational design of hepatitis C vaccines to hit the virus where it hurts.