Predicton of MHC Binding and Antigen Processing in T-Cell Epitope Identification

The immune system is a system of biological structures and processes within an organism that protects against diseases. When a host is infected by pathogens, physiological functions of the immune system trigger protective responses induced by parts of proteins known as epitopes. In the antigen presenting pathways, the transporter associated with antigen processing (TAP) delivers cleaved epitope peptides into the endoplasmic reticulum to bind to major histocompatibility complex (MHC) molecules. Then, the peptide-MHC complexes are presented to T-cell receptors on the surface of antigen-presenting cells. Despite recent technical advances, experimental determination of epitope binding remains time-consuming and labor-intensive. Thus, using computational approaches to extract immunological features from sequences have become highly important to understand relationship between hosts and pathogens and facilitate in silico epitope-based vaccine design. In this project, we propose a systematic approach to predict MHC binding affinity and analyze antigen processing events in T-cell epitope identification. Our prediction methods incorporate various physicochemical properties as feature representation, correspondence analysis as feature reduction, and compact set analysis and support vector machines as machine learning techniques. This study can help discover new vaccines and therapies for human immunodeficiency virus (HIV), malaria, tuberculosis, Ebola virus, and influenza. Moreover, the proposed immunological features can provide valuable insights into the nature of cancer, allergy, and autoimmune diseases. In the first year, we will aim at developing methods to predict MHC peptide binding and estimate binding affinity based on biological features both matrix-based and peptide based features. First, we analyze MHC binding data sets and generate novel matrix-based features including amino acid pair antigenicity and position-specific scoring matrix. Then, various physicochemical peptide-based features from AAindex database are reduced by correspondence analysis. Finally, reduced features are applied in a hybrid prediction method combined with compact sets and support vector machines, and immunological insights are drawn from our results with references to vaccine design. In the second year, we will focus on antigen processing pathway events that also play an important role for epitope prediction. Before antigenic peptides bind to MHC molecules, they must be cleaved and bind to TAP binders in the antigen presenting pathway. Thus, it is crucial to investigate the cleavage sites and TAP binding affinity. We will develop systems for predicting proteasomal cleavage sites and TAP binders. Finally, a systematic epitope prediction method will be proposed by integrating T-cell identification, MHC binding prediction, and antigen processing event analysis. In the coming two years, we will endeavor to develop useful bioinformatics tools and propose interpretable biological features that can be used collectively to assist immunologist in epitope-based vaccine design.