Figure 2

Strategy employed to assess the predictors’ performance. The scale bar represents the sequence of a theoretical protein of 120 amino acids. Dark blue rectangles represent one single epitope or a consensus of overlapping epitopes that were experimentally validated according to IEDB (Positive exp.); light blue rectangles represent one single non-immunogenic region or a consensus of overlapping non-immunogenic regions that were experimentally validated according to IEDB (Negative exp.); red rectangles represent the predicted epitopes from evaluated algorithms. For B cell prediction the predicted epitopes were considered true positive if they aligned with minimum coverage of 50% and 100% of identity with a Positive Exp. Region; for CD8+ T cell prediction the predicted epitopes were considered true positive if they aligned with minimum coverage of 87% and 100% of identity with a Positive Exp. region (Prediction 1, Prediction 2, Prediction 3 and Prediction 11). For B cell prediction the predicted epitopes were considered false positive if they aligned with minimum coverage of 50% and 100% of identity with a Negative Exp. Region; for CD8+ T cell prediction the predicted epitopes were considered false positive if they aligned with minimum coverage of 87% and 100% of identity with a Negative Exp. region (Prediction 6, Prediction 7, Prediction 8 and Prediction 13). Predictions were not considered during the analysis if they did not align with the parameters cited above (Prediction 4, Prediction 5, Prediction 9 and Prediction 10) or if they aligned with both Positive exp. and Negative exp. (Prediction 12).