Accurate Prediction of Repeatability of Peptide Identification from a Single Shotgun Proteomics Experiment

摘要

Due to instrumental, operational and environmental uncertainties, peptides and protein identifications from LC-MS/MS based proteomics experiments appear be stochastic although not completely random. Even repeated experiments of the same sample in the same platform (i.e. technical replicates) demonstrate significant variability [Tabb2010, Rudnick2009]. The variability greatly complicated the interpretation of results from the LC-MSMS experiments. In depth understanding of the sources and characteristics of such variability would be crucial for the quality control and subsequent applications of LC/MS/MS in biomedical sciences and clinical settings [Tabb2010, Rudnick2009]. In particular, the proteotypic peptides, which are repetitively identified in replicates experiments, can be used as candidates for constructing spectrum library [Craig2005] or for targeted proteomics [Fusaro2009].