Benchmarking dataset for optimizing differential expression analysis in label-free quantitative proteomic analysis

Abstract

The identification of accurate differentially expressed proteins (DEPs) is of paramount for proteomics analysis and biological interpretation. Other than the match between runs approach, various imputation and differential expression (DE) strategies have been developed to address the issue of missing values generated by the stochastic nature of data-dependent acquisition in MS-based proteomics¹. However, a comprehensive evaluation of the analysis pipelines’ performance in accurately identifying protein candidates is still lacking. To address this gap and benchmark the effectiveness of these methods, a proteomics dataset was generated, which encompasses proteins derived from human, yeast, and drosophila, each present in defined ratios². The data described here have been deposited to jPOST^3,4 with the identifier JPST001395.