Abstract
The application of in silico tools for the development of T-cell vaccines is crucial. Yet, due to myriad of polymorphisms of human T-lymphocytic antigen challenges, such therapeutic opportunities present unique roadblocks. There is an obvious advantage in using immunoinformatics (i.e., significantly decreasing cost related to laboratory expenses). A previous publication looked at random binding and nonbinding peptides in order to test the practicality of using such in silico tools to obtain possible immunogenic peptides. The present in silico study applied the same basic approaches to an applicable problem that was to identify promiscuous peptide vaccine candidates for hepatitis C virus (HCV) infection. The data sets used, included the proteins HCV E1, E2 and P7 as the binders (non-self antigens) and the GAD65 and ICA69, which have an association with diabetes, as non-binders (self-antigens). The in silico tools utilized were ProPred, MHC2PRED, and RANKPEP. The resulting differences were identifiable in each of the statistical parameters examined. Variations in the outcomes were evident by the dissimilarities found among the major indices of evaluation Sensitivity, Specificity, Accuracy, Positive Predictive Value (PPV), Negative Predictive Value (NPV) and Matthews's correlation coefficient (MCC) of the percentages of the predicted promiscuous peptides to HLA-DRB1*0101, *0301, and *0401. The conclusion from this study indicates that more work needs to be done in order to enhance the predictability of programs for the identification of peptide vaccine candidates for HCV. Such programs should not be solely relied upon without in vitro assay verification.