369861 Computational Design of Hepatitis C Vaccines
Hepatitis C virus (HCV) is a global scourge afflicting 170 million people, nearly 2.5% of the human population. Chronic HCV infection is responsible for a quarter of all cases of liver cirrhosis and cancer, and is the leading cause of liver transplants in the developed world. The gold standard treatment of daily injections and oral tablets costs more than $15,000 and has failure rates of 30-50%, making it only moderately effective and essentially unavailable in the developing world. Prophylactic vaccination offers the most realistic and cost effective hope of controlling the epidemic.
Despite 20 years of research, a vaccine remains elusive. Progress has been frustrated in large part because we do not know which components of the immune system to activate with a vaccine. By combining models from statistical physics with machine learning and data mining of viral sequence databases, we devised an approach to infer quantitative models for viral fitness that reveal viral “soft spots” that are vulnerable to vaccine-induced immune pressure.
We have applied our approach to the RNA-dependent RNA polymerase of HCV genotype 1a that is responsible for replication of the viral genome. The predictions of our model are in good accord with experimental measurements and clinical data, including: 1) in vitro replicative fitness measurements, 2) clinically documented escape mutations, 3) intra-host viral evolution revealed by longitudinal deep sequencing, and 4) protective HLA-associated cytotoxic T-cell responses.
We use our model to computationally screen candidate vaccine immonogens to perform in silico design of promising cytotoxic T-cell immunogens that exploit the identified viral vulnerabilities and have broad coverage within the target population. By reducing the time and expense of trial-and-error testing, our computational design platform can inform and accelerate the search for a vaccine to arrest the HCV epidemic.