Validation Tools for Predicted Linear B-Epitopes: Surface Flexibility

Abstract

Protein structural flexibility is important for catalysis, and binding. Flexibility has been predicted from amino acid sequence with a sliding window averaging technique and applied primarily to epitope search. Karplus, and Shultz divided amino acids in two categories, rigid and flexible. Then, proposed three flexibility scales for residues with no rigid neighbors, residues with one rigid neighbor, and residues for which both neighbors are rigid. To use flexibility scores for validation of candidate b-epitopes, 337,259 b-epitopes whose lengths are larger than seven amino acids are downloaded from iedb database, and 103,590 non epitopes are created randomly. When computed by Karplus-Shultz technique, b-epitopes achieved an average score of 1.0159, while the average score for non epitops is 0.9752. It is seen that the Karplus, and Shultz flexibility computation technique can be used as a validation tool with the criteria that the peptide with higher score, is more likely an epitope.