In silico assessment of somatic mutation generated T cell epitopes in various subsets of Acute Myeloid Leukemia

Abstract

Abstract The mechanisms of targeted-immunotherapy are still at the beginning stages for blood tumors. In the current era of big data and genomic sequencing, evolving bioinformatic tools are need for the validation and prediction of current and future immunotherapies. In particular, many neoepitope prediction programs do not consider germline context or the potential for co-occurrence of two or more somatic variants on the same mRNA transcript in the context of DNA-seq of complementary tumor and normal patient samples. Without consideration of these phenomena, existing approaches are likely to produce both false positive and false negative results, resulting in an inaccurate and incomplete picture of the cancer neoepitope landscape. In this study, we investigate the neoepitope landscape in largest-to-data dataset on acute myeloid leukemia (AML) from the Beat AML program (lead by OHSU), a cohort of 562 patients. Using a novel computational pipeline Neoepiscope (derived by Reid Thompson, MD PhD, OHSU) to chiefly address this issue for single nucleotide variants (SNVs) and insertions/deletions (indels), and herein illustrate how germline and somatic variant phasing affects neoepitope prediction. We hypothesize that in silico neoepitope prediction will accurately predict viable novel peptides that can be used for target immunotherapy in AML.