The in silico prediction of peptide binding affinities to MHC proteins is a very important first step in the process of epi-tope-based vaccine design and development. Five MHC class II binding prediction servers were ...The in silico prediction of peptide binding affinities to MHC proteins is a very important first step in the process of epi-tope-based vaccine design and development. Five MHC class II binding prediction servers were combined in different ways and the resulting performance of these combinations was evaluated using a test set, which consisted of 4540 known HLA-DRB1 binders. The five servers were: NetMHCIIpan, NetMHCII, ProPred, RANKPEP, and EpiTOP. The top 5% of the ranked predictions from each server were combined using union and intersection operators. The outputs were evaluated in terms of sensitivity and positive predictive value (PPV). The union operator showed high sensitivity (65-79%) and low PPVs (6-8%), while intersection outputs had low sensitivities (4-41%) yet significantly higher PPVs (14-31%). Thus there is a defining trade-off between sensitivity and PPV for each combination. The union of outputs from different servers brings more “noise” than “signal” to the resulting set of predicted binders. Conversely, selecting only commonly predicted binders increases the probability that an identified binder is a true binder.展开更多
文摘The in silico prediction of peptide binding affinities to MHC proteins is a very important first step in the process of epi-tope-based vaccine design and development. Five MHC class II binding prediction servers were combined in different ways and the resulting performance of these combinations was evaluated using a test set, which consisted of 4540 known HLA-DRB1 binders. The five servers were: NetMHCIIpan, NetMHCII, ProPred, RANKPEP, and EpiTOP. The top 5% of the ranked predictions from each server were combined using union and intersection operators. The outputs were evaluated in terms of sensitivity and positive predictive value (PPV). The union operator showed high sensitivity (65-79%) and low PPVs (6-8%), while intersection outputs had low sensitivities (4-41%) yet significantly higher PPVs (14-31%). Thus there is a defining trade-off between sensitivity and PPV for each combination. The union of outputs from different servers brings more “noise” than “signal” to the resulting set of predicted binders. Conversely, selecting only commonly predicted binders increases the probability that an identified binder is a true binder.
