AbstractThe immunopeptidome, representing the point of contact between somatic and T cells, is key for adaptive immunity. Each presented peptide holds an abundance of information not yet well understood. Up to now, the scientific focus has been the definition of pathogenic or tumor derived epitopes and the deconvolution of HLA peptide motifs of the entire immunopeptidome. Here we go one step further and assess the properties of individual peptides to identify defined HLA allotype-specific and frequently presented peptides. Such allotypic peptides represent a versatile tool to determine HLA allotypes or serve as internal standard for characterization of cancer antigens and differentially processed antigens. Finally, individual tissue- and dignity-specific antigens were defined, and the latter were successfully implemented for molecular tumor testing.Using mass spectrometry based immunopeptidomics a database was generated consisting of ∼900 HLA-typed samples. The identified allotypic peptides enabled a HLA class I allotype determination, which was 95% correct in our in-house dataset and 98% in an external dataset. These abundant peptides were implemented as internal standard for a semi-quantitative investigation of established tumor antigens and antigens processed differentially in malignant and benign tissue. Defined dignity-specific antigens allowed a 87% correct tumor detection across numerous tumor types.In summary, we describe a machine learning approach for mining immunopeptidomic data in order to develop a classification method, allowing to differentiate HLA class I-allotypes of a sample or distinguish between healthy and malignant state of tissues. Furthermore, based on this method, we developed a procedure for the validation of tumor exclusive antigens. Our results support classification of immunopeptidomic data sets using machine learning and highlight their potential utility for biomarker development.
Front cover: High-sensitivity HLA class I peptidome analysis enables a precise definition of peptide motifs and the identification of peptides from cell lines and patients’ sera