Colander: a probability-based support vector machine algorithm for automatic screening for CID spectra of phosphopeptides prior to database search

We developed a probability-based machine-learning program, Colander, to identify tandem mass spectra that are highly likely to represent phosphopeptides prior to database search. We identified statistically significant diagnostic features of phosphopeptide tandem mass spectra based on ion trap CID MS/MS experiments. Statistics for the features are calculated from 376 validated phosphopeptide spectra and 376 non-phosphopeptide spectra. A probability-based support vector machine (SVM) program, Colander, was then trained on five selected features. Datasets were assembled both from LC/LC-MS/MS analyses of large-scale phosphopeptide enrichments from proteolyzed cells, tissues and synthetic phosphopeptides. These datasets were used to evaluate the capability of Colander to select pS/pT-containing phosphopeptide tandem mass spectra. When applied to unknown tandem mass spectra, Colander can routinely remove 80% of tandem mass spectra while retaining 95% of phosphopeptide tandem mass spectra. The program significantly reduced computational time spent on database search by 60% to 90%. Furthermore, pre-filtering tandem mass spectra representing phosphopeptides can increase the number of phosphopeptide identifications under a pre-defined false positive rate.