Fully unsupervised identification of HLA-I motifs

The precise identification of Human Leukocyte Antigen class I (HLA-I) binding motifs plays a central role in our ability to understand and predict (neo-)antigen presentation in infectious diseases and cancer. Here, by exploiting co-occurrence of HLA-I alleles across publicly available as well as ten newly generated high quality HLA peptidomics datasets, we show that we can rapidly and accurately identify HLA-I binding motifs and map them to their corresponding alleles without any a priori knowledge of HLA-I binding specificity. This fully unsupervised approach uncovers new motifs for several alleles without known ligands and significantly improves neo-epitope predictions in three melanoma patients.