Extensive and accurate benchmarking of DIA acquisition methods and software tools using a complex proteomic standard

Abstract

In the past few years, LC-MS/MS in DIA mode has become a strategy of choice for deep coverage of complex proteomes and accurate quantification of low abundant species. However, there is still no consensus in the literature on the best acquisition parameters and processing tools to use. We present here the largest benchmark of DIA proteomic workflows on Orbitrap instruments ever published. Using a complex proteomic standard, we tested 36 workflows including 4 different acquisition schemes and 6 different software tools (DIA-NN, DIA-Umpire, OpenSWATH, Scaffold-DIA, Skyline and Spectronaut-Pulsar-X) with or without the use of a spectral library. By reporting the protein identifications, linearity, reproducibility, as well as sensitivity and specificity in 28 pairwise comparisons, we propose here a guide book for the users to choose the best DIA strategy for their data. Moreover, our data has been deposited on ProteomeXchange for testing or developing new DIA processing tools.

DIA proteomics data from a UPS1-spiked E.coli protein mixture processed with six software tools

“…pdresults file of Proteome Discoverer. Detailed parameters used to generate both spectral libraries are listed in the supplementary table S1 of the Gotti et al article [1] .…”

“…The dataset provided in this article has been initially generated with the aim to benchmark DIA acquisition methods and software tools [1] . As shown on Fig.…”

“…Fragments: 4; Max. Fragments: 6; Peptide charge: 2+ to 4+; Other settings (see Gotti et al Supp Table S1 [1] ) 1% FDR or q-value < 0.01 at precursor, peptide and/or protein level (see Gotti et al.Supp Table S1 [1] ) DIA-Umpire [5] 2.0 .mzXML yes (DIA spectral library from DIA files) no OpenSWATH (diaproteomics) [6 , 7] 1.1.0 .mzML no yes (Skyline .blib exported to .tsv) ScaffoldDIA (Proteome Software, Inc.) [8] 2.1.0 none yes (in silico Prosit library) yes (Skyline .blib) Skyline [2] 20.2.0.343 .mzML no yes (Skyline .blib) Spectronaut (Biognosys AG) [3] 14.10.20122 none yes ('directDIA' mode) yes (Spectronaut library) …”

DIAproteomics: A multi-functional data analysis pipeline for data-independent-acquisition proteomics and peptidomics

“…An overview over available parameters and a short description is provided at https://nfco.re/diaproteomics. The default parametrization has been benchmarked multiple times in the past 23,37 . It involves spectral library assay generation with the six most intense b-and y-ion transitions falling into the precursor mass range of 400 to 1200 m/z and a fragment mass range of 350 to 2000 m/z.…”

In-Depth Benchmarking of DIA-type Proteomics Data Analysis Strategies Using a Large-Scale Benchmark Dataset Comprising Inter-Patient Heterogeneity

“…sequence and concentration) are added to 'background' peptides with likewise known properties. To mimic the complexity encountered in realistic settings often different organisms are added in combination to create benchmark datasets 4 . These benchmark datasets are valuable tools for controlling and optimizing different aspects of data acquisition and analysis, including LC-MS/MS parameters, library generation, analysis software parameters, data preprocessing and statistical analysis for detecting differentially abundant proteins.…”