It started with a simple question: can we use the mutations in a tumor to immunize against itself. We looked at this very carefully in melanomas, fibrosarcomas, ovarian and prostate cancers, in mice and humans. After we have tested over 1000 peptides containing the mutations that we identified in the tumors, the answer is yes. It’s an exciting finding that those mutations are quite unique to individual tumors, and in the mouse models, epitopes containing those tumor rejecting mutations have weak binding affinity to MHC I molecules. This findings not only set the foundation for the development of customized cancer vaccines, but also help to address some fundamental immunological questions like how tumor antigens are recognized by T-cells.