Vaccines have saved more lives than any other medical intervention throughout human history by preventing the spread of infectious diseases. However, despite several decades of research, there is no effective vaccine against fast evolving viruses such as the human immunodeficiency virus (HIV) and the hepatitis C virus (HCV). A confounding factor in the development of a HIV or HCV vaccine is that these viruses have a unique ability to make a lot of mutations in their genetic code. This enables them to escape the human immune system while retaining their ability to propagate infection. For developing a vaccine against such viruses, scientists are developing novel strategies which seek to target specific parts of the virus that are most vulnerable (i.e., where it is difficult for the virus to survive mutations) in order to induce a focused and potentially effective immune response. To determine the existence and location of such parts of HIV and HCV, initial studies have leveraged recently-available sequence data for these viruses, and looked for those positions in the genome for which the frequency of mutation was lowest. Unfortunately, vaccines based on such first-order statistics have not enjoyed much success, and there is increasing evidence suggesting that interactions between mutations is also important and must be considered when designing an effective vaccine against HIV and HCV. It is almost impossible to determine effects of interactions between all mutations experimentally as it requires performing billions of experiments. In this article, we explain how by leveraging virus sequence data, mutational interactions can be estimated using statistical techniques and incorporated in designing novel and potentially effective vaccine strategies against such fast-evolving viruses.