Abstract 1895: caAtlas: An immunopeptidome atlas of human cancer

Abstract

Abstract At the core of immunotherapy is the T cell recognition of tumor antigens. Thus, comprehensive characterization of tumor antigens is essential for the design of effective and safe cancer immunotherapy strategies. Mass spectrometry (MS)-based immunopeptidomics allows high-throughput, direct identification of major histocompatibility complex (MHC)-bound peptides in vivo. Here, we collected 42 published human immunopeptidomic datasets, which include 327 cancer samples covering nine cancer types and 216 non-cancerous samples. A systematic, quality-controlled analysis of all 51.6 million MS/MS spectra using an open search algorithm identified 364,283 peptides, including 11,647 modified peptides associated with 80 types of modifications. Unmodified peptides showed high levels of predicted binding affinities by NetMHCpan; however, the predicted binding affinities for the modified peptides were much lower, suggesting a deficiency of existing computational tools in making predictions for modified peptides. Analysis of the 416 phosphorylated and 1,239 acetylated HLA class I peptides revealed distinct sequence motifs, and the pattern for the phosphopeptides was consistent with previous reports. Further interrogation of all identified peptides revealed evidence for tumor specific-presentation of hundreds of known and novel cancer-testis antigens as well as tumor associated antigens. We also found non-tumor-specific presentation of 44 previously annotated cancer-testis antigens. Applying NeoQuery, a newly developed computational pipeline, to our dataset identified immunopeptidomic evidence for more than 1,000 somatic mutation-derived neoantigens, and a subset of them were further validated through a traditional MS/MS peptide identification pipeline. We make all these data together with annotated MS/MS spectra supporting each individual antigen identification in a publicly accessible, easily browsable web portal named cancer antigen atlas (caAtlas, http://www.zhang-lab.org/caatlas/). caAtlas provides a fundamental resource for the selection and prioritization of MHC-bound peptides for immunogenicity testing and cancer immunotherapy development. Citation Format: Xinpei Yi, Yuxing Liao, Bo Wen, Kai Li, Yongchao Dou, Sara R. Savage, Bing Zhang. caAtlas: An immunopeptidome atlas of human cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1895.