Optimization of the Use of Consensus Methods for the Detection and Putative Identification of Peptides via Mass-spectrometry Using Protein Standard Mixtures

Abstract

Correct identification of peptides and proteins in complex biological samples from proteomic mass-spectra is a challenging problem in bioinformatics. The sensitivity and specificity of identification algorithms depend on underlying scoring methods, some being more sensitive, and others more specific. For high-throughput, automated peptide identification, control over the algorithms' performance in terms of trade-off between sensitivity and specificity is desirable. Combinations of algorithms, called 'consensus methods', have been shown to provide more accurate results than individual algorithms. However, due to the proliferation of algorithms and their varied internal settings, a systematic understanding of relative performance of individual and consensus methods are lacking. We performed an in-depth analysis of various approaches to consensus scoring using known protein mixtures, and evaluated the performance of 2310 settings generated from consensus of three different search algorithms: Mascot, Sequest, and X!Tandem. Our findings indicate that the union of Mascot, Sequest, and X!Tandem performed well (considering overall accuracy), and methods using 80-99.9% protein probability and/or minimum 2 peptides and/or 0-50% minimum peptide probability for protein identification performed better (on average) among all consensus methods tested in terms of overall accuracy. The results also suggest method selection strategies to provide direct control over sensitivity and specificity.

Discovery of Mouse Spleen Signaling Responses to Anthrax using Label-Free Quantitative Phosphoproteomics via Mass Spectrometry

“…Compared with using either Mascot or SEQUEST alone, using both search engines (i.e. "consensus scoring") resulted in roughly triple the number of confidently identified phosphopeptides, in agreement with previous studies (51)(52)(53)(54)(55).…”

“…This is not surprising, as the use of multiple search engines (i.e. "consensus scoring") has been shown to significantly improve peptide identification confidence (51)(52)(53)(54)(55). Additionally, a short Perl script was written (supplemental material) to remove tentative neutral-loss precursor ion peaks from the fragment ion spectra .dta files prior to the Mascot and SEQUEST analyses, and this significantly increased the number of identified phospho-peptides at the same FDR.…”

Effect of cleavage enzyme, search algorithm and decoy database on mass spectrometric identification of wheat gluten proteins

“…Recently an evaluation of search algorithm combinations (Dagda et al, 2010;Sultana et al, 2009) demonstrated that, in general, the union of Mascot and X!Tandem provided high specificity and reasonable sensitivity. Other investigators have utilized chemical modification (Rehulkova et al, 2009), different enzymes (Swaney et al, 2010) or multiple search engines to increase protein coverage.…”

Combining Results of Multiple Search Engines in Proteomics

“…Many algorithms and software tools have emerged that combine search results (21)(22)(23)(24)(25)(26)(27)(28), each demonstrating an improved final result over any individual search engine alone. Such improved results come at the cost of the increased complexity of managing multiple searches in an analysis pipeline, as well as a several-fold increase in computational time in what is already the most computationally expensive step.…”