Affiliations | Department of Oncology, Ludwig Institute for Cancer Research, University of Lausanne, 1066 Epalinges, Switzerland; david.gfeller@unil.ch; Swiss Institute of Bioinformatics, 1015 Lausanne, Switzerland; Department of Oncology, Ludwig Institute for Cancer Research, University Hospital of Lausanne, 1011 Lausanne, Switzerland; and; Service of Neurosurgery, University Hospital of Lausanne, 1011 Lausanne, Switzerland. | |
Abstract | HLA-I molecules bind short peptides and present them for recognition by CD8<sup>+</sup> T cells. The length of HLA-I ligands typically ranges from 8 to 12 aa, but variability is observed across different HLA-I alleles. In this study we collected recent in-depth HLA peptidomics data, including 12 newly generated HLA peptidomes (31,896 unique peptides) from human meningioma samples, to analyze the peptide length distribution and multiple specificity across 84 different HLA-I alleles. We observed a clear clustering of HLA-I alleles with distinct peptide length distributions, which enabled us to study the structural basis of peptide length distributions and predict peptide length distributions from HLA-I sequences. We further identified multiple specificity in several HLA-I molecules and validated these observations with binding assays. Explicitly modeling peptide length distribution and multiple specificity improved predictions of naturally presented HLA-I ligands, as demonstrated in an independent benchmarking based on the new human meningioma samples. | |