Repertoire forecast of MHC class I binding peptides with peptide libraries

Understanding the repertoire of MHC binding peptides is a pivotal step towards looking into the genetic forces that keep the MHC molecules polymorphic and polygenic. We have attempted to forecast the repertoire of MHC class I binding peptides by means of positional scanning using peptide libraries. Position independent assessment of the binding was useful for roughly estimating the binding capacity of peptides, but the prediction was clearly limited. Namely, for approximately 80% of randomly chosen peptides, MHC binding capacity could be predicted within an order of magnitude in Kd values. When pairwise comparisons were made between different MHC class I molecules, the degree of repertoire overlap was 2 to 5%. The results support the leading hypothesis that the necessity to increase repertoire size acts as a driving force for diversification of peptide binding specificities in the MHC class I molecules.