P17-M Improving Sensitivity by Combining Results from Multiple MS/MS Search Methodologies with the Scaffold Computer Algorithm

摘要

Database-searching programs generally identify only a fraction of the spectra acquired in a standard LC/MS/MS study of digested proteins. Subtle variations in database-searching algorithms of MS/MS spectra have been known to provide different identification results. To leverage this variation, we developed Scaffold to probabilistically combine the results of multiple search engines, including Sequest, Mascot, and X!Tandem. Here we present a “tell all” explanation of the specific methodology behind Scaffold that converts scores into search engine independent peptide probabilities. These probabilities can be readily combined across search engines using Bayesian rules and the Expectation Maximization learning algorithm. We demonstrate how we normally gain 20% to 100% more highly confident (>95%) MS/MS spectrum identifications with each additional search engine, which is primarily due to increased confidence in low-scoring matches. We also show that this method works reliably across a variety of search engines and instrumentation platforms without re-tuning. >Figure 1