The Research Of B-cell Epitope Prediction Algorithm Based On Phage Display

The research of B cell epitope has always been an important area of immunology. The traditional way to map the B cell epitope is no longer suitable for current need. In recent decades, the prediction of B cell epitope is developing very fast,including the prediction of conformational B-cell epitopes based on phage display experiment. A set of mimotope from the phage display experiment with a given antibody as the target and the three-dimensional structure of corresponding antigen are required for this kind of B cell epitope prediction. The basic idea is that mimotopes are similar to the relevant B-cell epitope.Based on the physicochemical properties of amino acids, we proposed MIMOX2, a new conformational B-cell epitope prediction algorithm based on phage display. This algorithm divides the surface of antigen into overlapping patches and evaluates their similarities to mimotopes. In order to facilitate the use of this algorithm, we supply the web service which is available at:http://immunet.cn/mimox/cgi-bin/MIMOX2.pl.As the field of such kind of B-cell epitope prediction is lack of standard evaluation dataset, we constructed a standard test data set called WXY, which is based on the protein data bank(PDB) and the world’s largest phage display database MimoDB. WXY has 70 sets of test data, from four categories, i.e. Antibody-Antigen complex, Ligand-Receptor complex, Other Protein-Protein complex and Small Molecule-Protein complex.Based on the WXY data set, we evaluated MIMOX2, PepSurf, Mapitope, and Pepitope systematically with Mathew correlation coefficient, Positive Predictive Value, Sensitivity,and specificity. In many cases, MIMOX2 showed better performance, suggesting it can complement relevant tools.