Ventsislav Yordanov, Ivan Dimitrov, Irini Doytchinova
The recognition of a foreign protein (antigen) by host B and/or T cells and the induction of immune response have a key role for the human immune system. The peptide binding to MHC proteins is a prerequisite for B or T-cell recognition. Determining the peptide-binding preferences exhibited by the MHC extensive set of alleles requires an enormous experimental work. A more rational approach is the development of bioinformatics prediction methodologies. Among others, proteochemometrics (PCM) is quite suitable for MHC binding prediction as it models simultaneously the bioactivity of multiple ligands against multiple protein targets. Handling multiple targets is the key to "ligand polypharmacology" and the development of multi-target drugs. The few applications of PCM for MHC binding prediction showed that the developed models have excellent predictive ability.