Identification and Ranking of Recurrent Neo-Epitopes in Cancer

Abstract

Neo-epitopes are emerging as attractive targets for cancer immunotherapy and new strategies for rapid identification of relevant candidates have become a priority. We propose a method for in silico selection of candidates which have a high potential for neo-antigen generation and are likely to appear in multiple patients. This is achieved by carefully screening 33 TCGA data sets for recurrent somatic amino acid exchanges and, for the 1,055 resulting recurrent variants, applying MHC class I binding prediction algorithms. A preliminary confirmation of epitope binding and recognition by CD8 T cells has been carried out for a couple of candidates in humanized mice. Recurrent neo-epitopes may be suitable to supplement existing personalized T cell treatment approaches with precision treatment options.