The peptide identifications are improved afterwards by filtering out all peptides which have a large deviation between observed and predicted retention time.Performance of Peptide Sample Fractionation PredictionTo be able to compare our results with existing methods, we used the same dataset and the same setup as Oh et al. [11].