Multiple regression analysis suggests motifs for class II MHC binding

Multiple regression analysis is utilized to predict peptides that bind major histocompatibility complex (MHC) class II molecules of the d haplotype. The predictor variables, obtained from the primary sequence of protein fragments, include pairs of amino acid residues separated by zero to seven intervening residues. The resulting models demonstrate strong predictive validity, and have interpretations in accordance with recent discoveries regarding the binding mechanisms of class I MHC molecules.