Data extraction from proteomics raw data: an evaluation of nine tandem MS tools using a large Orbitrap data set.

Area of Science:

• Proteomics
• Mass Spectrometry
• Bioinformatics

Background:

• Shot-gun proteomics relies on processing raw tandem mass spectrometry (MS) data into peak lists for database searching.
• Numerous tandem MS extraction tools exist, but a systematic comparison has been lacking.

Purpose of the Study:

• To systematically compare the performance of nine different tandem MS extraction tools.
• To evaluate variations in extracted data, peptide identification, and tool agreement.

Main Methods:

• Utilized over 400,000 tandem MS spectra from an Orbitrap Velos instrument.
• Compared nine freely available and commercial tandem MS extraction tools.
• Assessed metrics including MS/MS events, fragment ion data, matches, precursor mass accuracy, and inter-tool agreement.

Main Results:

• Processing the same raw data with different tools resulted in low overlap of identified peptides.
• Significant differences were observed in precursor charge states, precursor and fragment ion masses.
• Highly confident peptide identifications from one tool were not always matched by others.
• A bias towards lower charge states was noted when extracting data from high-resolution raw data without deconvolution.

Conclusions:

• Tandem MS extraction tools exhibit considerable variability, impacting peptide identification confidence and consistency.
• Understanding these differences is essential for optimizing parameter settings and tool selection.
• Choosing the right extraction tool is critical for reliable and reproducible proteomics data analysis.