Gregory Hart: CSE 2015 Fellow

Physics

In Silico Vaccine Design by Inverse Inference of Viral Fitness Landscapes

As the price and difficulty of sequencing DNA has dropped, sequence databases have swelled. I am trying to leverage this data to create a computational platform to design and prescreen vaccine candidates for highly variable viruses.

Using sequences from hepatitis C virus databases, I have been able to find empirical fitness landscapes for several viral proteins. The fitness landscape reveals how the fitness of the virus changes with different mutations. This allows me to identify the price the virus pays to escape various immune responses, and thus, which responses do the most damage to the virus. This is the basis of a simple prescreening process that allows millions of vaccine candidates to be reduced to a few 100 optimal candidates.

Currently, I am working on integrating the fitness landscapes into population dynamics and immunological simulations. Such a model will reveal how the virus evolves under immune pressure. This allows me to test the vaccine candidates and answers questions such as: How quickly does the immune response suppress the virus? Is the virus able to effectively escape the immune response? How long does it take to escape, and what escape pathways does it use? Can these pathways be blocked? How robust is the vaccine to different infecting strains? With this knowledge the vaccine candidates can be further optimized before expensive and time-consuming pre-clinical and clinical testing.

For a brief research video: https://www.youtube.com/watch?v=IrRNPNNX9UY&feature=youtu.be

Image of Hepatitus C virus copyright Russell Kightley, used with permission.

Back to 2015-2016 Fellows