QSAR and the Prediction of T-Cell Epitopes

Abstract

Quantitative structure – activity relationships (QSAR) is a well established ligand-based approach to drug design. It correlates changes in the chemical structure of a series of compounds with changes in their biological activities. Peptides of equal length which bind to a certain protein are an excellent target for QSAR. In the present review, we summarize our experience in QSAR studies of peptides acting as T-cell epitopes. T-cell epitopes are protein fragments presented on the cell surface which afford the immune system the opportunity to detect and respond to both intracellular and extracellular pathogens. Epitope-based vaccines are a new generation of vaccines with lower side effects. The process of antigen presentation, which includes proteasome cleavage, TAP and MHC binding, has been modeled and analyzed by QSAR. Derived QSAR models are highly predictive, allowing us to design and test in vitro MHC superbinders. All models have been implemented in servers for in silico prediction of MHC binders and T-cell epitopes. In practice, better initial in silico prediction leads to improved subsequent experimental research on epitope-based vaccines. Keywords: Proteasome, TAP, MHC class I, MHC class II, additive method