Comprehensive Evaluation and Optimization of the Data-Dependent LC–MS/MS Workflow for Deep Proteome Profiling

Abstract

Data-dependent liquid chromatography–tandem mass spectrometry (LC–MS/MS) is widely used in proteomic analyses. A well-performed LC–MS/MS workflow, which involves multiple procedures and interdependent metrics, is a prerequisite for deep proteome profiling. Researchers have previously evaluated LC–MS/MS performance mainly based on the number of identified peptides and proteins. However, this is not a comprehensive approach. This motivates us to develop MSRefine, which aims to evaluate and optimize the performance of the LC–MS/MS workflow for data-dependent acquisition (DDA) proteomics. It extracts 47 kinds of metrics, scores the metrics, and reports visual results, assisting users in evaluating the workflow, locating problems, and providing optimizing strategies. In this study, we compared and analyzed multiple pairs of datasets spanning different samples, methods, and instruments and demonstrated that the comprehensive visual metrics and scores in MSRefine enable us to evaluate the performance of the various experiments and provide optimal strategies for the identification of more peptides and proteins.

Observations from the Proteomics Bench

“…Bioinformatics steps should, however, not be misunderstood as post-experiment salvation of suboptimal study design or sample quality; they are rather inherent steps of the workflow that can only properly deal with data obtained in established standardized experiments of a guaranteed quality level. Many authors have worked on the improvement of MS workflows (for examples, see references [20][21][22]). Here, we discuss the way in which we control our experiments.…”