The advancement in fast DNA sequencing technologies has opened up new opportunities to explore a diverse set of questions in biomedical research. In this paper, we review a general method which utilizes the available sequence data to determine potential weaknesses in highly mutable viruses, and which has shown promise in the design of vaccines. A key computational part of this method employs concepts from random matrix theory to obtain a robust estimate of a large covariance matrix. We apply this general method on hepatitis C virus as an example, and verify its usefulness by linking with the existing experimental and structural data. © 2014 IEEE.