Empirical, AI, and QSAR approaches to peptide-MHC binding prediction
Hattotuwagama CK., Guan P., Davies M., Taylor DJ., Walshe V., Hemsley SL., Toseland C., Doytchinova IA., Borrow P., Flower DR.
© 2007 Springer Science-hBusiness Media, LLC. All rights reserved. Immunoinformatics is facilitating important change within immunology and is helping it to engage more completely with the dynamic post-genomic revolution sweeping through bioscience. Historically, predicting the specificity of peptide Major Histocompatibility Complex (MHC) interactions has been the major contribution made by bioinformatics disciplines to research in immunology and the vaccinology. This will be the focus of the current chapter. Initially, we will review some background to this problem, such as the thermodynamics of peptide binding and the known constraints on peptide selectivity by the MHC. We will then review artificial intelligence and machine learning approaches to the prediction problem. Finally, we will outline our own contribution to this field: the application of QSAR techniques to the prediction of peptide-MHC binding.