Neoantigens are peptides on the surface of cancer cells presented to the immune system. Multiple novel therapeutic approaches involve the administration of neoantigens to trigger immunity-induced tumor regression. Identification of neoantigens includes a personalized approach consisting of detailed analyses of the sequenced tumor tissue and its comparison with wild type to identify somatic mutations. Alternated peptides are translated from nucleotides around somatic mutations and their binding affinity and immunogenicity need to be further evaluated. Still, the entire bioinformatics analysis is very complex, and accurate prediction of the neoantigen candidates represents a true challenge. Here, we present the novel, integral bioinformatic analysis workflow for neoantigen discovery, denoted INAEME (Integral Neoantigen Analysis with Entirety of Mutational Events). The workflow performs integral processing of an individual’s DNA tumor-normal and RNA tumor raw reads to output prioritized neoantigen candidates. Our evaluation analysis includes a wide scope of mutational events so far not considered in the existing solutions, including phasing of variants, influence of both somatic and germline variants, positions of all transcripts, neighboring variants, and frameshifts. The influence of each mutational event on the accuracy of predicted neoantigen candidates is tested across 300 TCGA samples from multiple cancer types. The obtained results have demonstrated the significance of considering the entirety of mutational events to obtain an accurate set of strong neoantigen candidates for cancer immunotherapy targets or vaccines. The adaption of the described methods in the bioinformatics analysis minimizes the existence of false positives which are only later discovered in a laboratory environment using expensive methods such as mass spectrometry or microscopy.
