Predicting MHC-I ligands across alleles and species: How far can we go?

Abstract

CD8+ T-cell activation is initiated by the recognition of epitopes presented on class I major histocompatibility complex (MHC-I) molecules. Identifying such epitopes is useful for molecular understanding of cellular immune responses and can guide the development of personalized vaccines for various diseases including cancer. Here, we capitalize on high-quality MHC-I peptidomics data available from different species and an expanded architecture of our MHC-I ligand predictor (MixMHCpred) to carefully explore how much predictions can be extrapolated to MHC-I alleles without known ligands. Our results reveal high prediction accuracy for most MHC-I alleles in human and in laboratory mouse strains, but significantly lower accuracy in other species. Our work further outlines some of the molecular determinants of MHC-I ligand predictions accuracy across alleles and species. Robust benchmarking on external data shows that the pan-allele version of MixMHCpred (v3.0) outperforms other state-of-the-art MHC-I ligand predictors and can be used for CD8+ T-cell epitope predictions.