Statistical fingerprints of immune selection of tumor cells

The adaptive immune system plays an evolutionary arms race against pathogens and tumor cells. The constant generation of T cells with novel T cell receptors (TCRs) creates pressure on viral epitopes and cancer neoantigens. In cancer, this process can be analyzed by following both the neoantigens and the immune repertoire through sequencing. On the immune side, it is possible to sequence TCRs from T cells in blood and tumor samples, getting a snapshot of the changing immune system. Selection effects would affect the T cell repertoire diversity, cluster TCRs in amino acid sequence space and change their amino acid statistics, with these changes reflected in TCR samples. I will describe computational methods that can be used to infer immune selection effects from data, based on quantitative analysis of TCRs sequences. We see that changes in the diversity and focusing of the repertoire are indicative of the affect the immune system has on the tumor cells.