5. Conclusions
T cell epitopes of seven soybean allergens were predicted by using the“MHC-II Binding Predictions” tools in the IEDB database, and 138 T cell epitopes were obtained by the further evaluation of the ability to induce the production of IL4 and allergenicity by bioinformatics tools, and 5 predicted fragments of “YIKDVFRVIPSEVLS” (aa 477-491), “KDVFRVIPSEVLSNS” (aa 479-493), “DVFRVIPSEVLSNSY” (aa 480-494) in P04347 protein and “AKADALFKAIEAYLL” (aa 138-152), “ADALFKAIEAYLLAH” (aa 140-154) in P26987 protein were considered as the most possible epitope candidates. Based on the amino acid composition analysis and random forest model, hydrophobic residues (such as phenylalanine, isoleucine, and valine) at positions p1 and p2 and positively charged amino acid residues (such as lysine and histidine) in positions p13 would have a good contribution to the allergenicity. Furthermore, most of the epitopes of T cells could be hydrolyzed by pepsin into small molecular peptides (<12 aa), and anti-digestive epitope regions contained more isoleucine (I), valine (V), serine (S), asparagine (N), and glutamine (Q). In the current studies, bioinformatics strategies provide an efficient method for epitope prediction and analysis, and the results may aid to study the mechanism of soybean sensitization. Additionally, epitopes reported here are expected to apply in immunotherapy designing and hypoallergenic soybean products developing in the food industry.