Knowledge-based structure prediction of MHC class I bound peptides: a study of 23 complexes

Background: The binding of T-cell antigenic peptides to MHC molecules is a prerequisite for their immunogenicity. The ability to identify binding pept ides based on the protein sequence is of great importance to the rational d esign of peptide vaccines. As the requirements for peptide binding cannot b e fully explained by the peptide sequence per se, structural considerations should be taken into account and are expected to improve predictive algori thms. The first step in such an algorithm requires accurate and fast modeli ng of the peptide structure in the MHC-binding groove. Results: We have used 23 solved peptide-MHC class I complexes as a source o f structural information in the development of a modeling algorithm. The pe ptide backbones and MHC structures were used as the templates for predictio n. Sidechain conformations were built based on a rotamer library, using the 'dead end elimination' approach. A simple energy function selects the favo rable combination of rotamers for a given sequence. It further selects the correct backbone structure from a limited library. The influence of differe nt parameters on the prediction quality was assessed. With a specific rotam er library that incorporates information from the peptide sidechains in the solved complexes, the algorithm correctly identifies 85% (92%) of all (bur ied) sidechains and selects the correct backbones. Under cross-validation, 70% (78%) of all (buried) residues are correctly predicted and most of all backbones. The interaction between peptide sidechains has a negligible effe ct on the prediction quality. Conclusions: The structure of the peptide sidechains follows from the inter actions with the MHC and the peptide backbone, as the prediction is hardly influenced by sidechain interactions. The proposed methodology was able to select the correct backbone from a limited set. The impairment in performan ce under cross-validation suggests that, currently, the specific rotamer li brary is not satisfactorily representative. The predictions might improve w ith an increase in the data.