TIDD: tool-independent and data-dependent machine learning for peptide identification

Abstract

Background In shotgun proteomics, database search engines have been developed to assign peptides to tandem mass (MS/MS) spectra and at the same time post-processing (or rescoring) approaches over the search results have been proposed to increase the number of confident peptide identifications. The most popular post-processing approaches such as Percolator and PeptideProphet have improved rates of peptide identifications by combining multiple scores from database search engines while applying machine learning techniques. Existing post-processing approaches, however, are limited when dealing with results from new search engines because their features for machine learning must be optimized specifically for each search engine. Results We propose a universal post-processing tool, called TIDD, which supports confident peptide identifications regardless of the search engine adopted. TIDD can work for any (including newly developed) search engines because it calculates universal features that assess peptide-spectrum match quality while it allows additional features provided by search engines (or users) as well. Even though it relies on universal features independent of search tools, TIDD showed similar or better performance than Percolator in terms of peptide identification. TIDD identified 10.23–38.95% more PSMs than target-decoy estimation for MSFragger, which is not supported by Percolator. TIDD offers an easy-to-use simple graphical user interface for user convenience. Conclusions TIDD successfully eliminated the requirement for an optimal feature engineering per database search tool, and thus, can be applied directly to any database search results including newly developed ones.

Current progress and critical challenges to overcome in the bioinformatics of mass spectrometry-based metaproteomics

“…There are several approaches to address the problem of reduced identified peptides in target-decoy searches using large-scale databases. These can be broadly divided into two types, step-by-step narrowing searches [67] , [68] , [69] , [70] , [71] and post-processing approaches such as rescoring [72] , [73] , [74] , [75] . The former is an attempt to solve the problem of target-decoy search by selecting important sequences by repeating the database search.…”

“…Here, we introduce a method for solving the problems of false positives and false negatives by re-evaluating the scores obtained as a result of a conventional target-decoy search. Post-scoring includes methods such as percolator [72] , peptide prophet [73] , Tool-Independent and Data-Dependent PSM rescoring (TIDD) [74] , and BH-FDR (Benjamini-Hochberg framework) [75] . Percolator uses the matching score as a result of the conventional target-decoy search ( e.g.…”

“…The data used for machine learning depend on and must be optimized for each search engine. The TIDD proposed by Li et al [74] uses SVM to optimize scores similar to the percolator but introduces parameters that are independent of search engines. This makes it possible to reliably obtain PSMs regardless of the search engine used.…”