The rule based classification models for MHC binding prediction and identification of the most relevant physicochemical properties for the individual allele

Abstract

Binding of proteolyzed fragments of proteins to MHC molecules is essential and the most selective step that determines T-cell epitopes. Therefore, the prediction of MHC-peptide binding is principal for anticipating potential T cell epitopes and is of immense relevance in vaccine design. Despite numerous methods for predicting MHC binding ligands, there still exist limitations that affect the reliability of a prevailing number of methods. Certain important methods based on physicochemical properties have very low reported accuracy. The aim of this paper is to present a new approach of extracting the most important physicochemical properties that influence the classification of MHC-binding ligands. In this study, we have developed rule based classification models which take into account the physicochemical properties of amino acids and their frequencies. The models use k-means clustering technique for extracting the relevant physicochemical properties. The results of the study indicate that the physicochemical properties of amino acids contribute significantly to the peptide-binding and that the different alleles are characterized by a different set of the physicochemical properties.