Simulated-to-real benchmarking of acquisition methods in untargeted metabolomics

Area of Science:

• Metabolomics
• Analytical Chemistry
• Computational Chemistry

Background:

• Data-Dependent Acquisition (DDA) and Data-Independent Acquisition (DIA) are standard for untargeted metabolomics LC-MS/MS.
• Systematic comparisons of DDA and DIA MS/MS spectral annotation are limited by experimental costs and lack of ground truth.

Purpose of the Study:

• To perform a systematic in silico comparison of DDA and DIA acquisition modes for untargeted metabolomics.
• To validate simulation results with real-world mass spectrometry data.

Main Methods:

• Extended the Virtual Metabolomics Mass Spectrometer (ViMMS) framework with a DIA module for in silico simulations.
• Compared DDA and DIA performance based on varying numbers of co-eluting ions.
• Validated simulation findings on an actual mass spectrometer.

Main Results:

• Acquisition mode performance is highly dependent on the number of co-eluting ions.
• DIA outperforms DDA when co-eluting ion numbers are low.
• DDA shows an advantage over DIA when co-eluting ion numbers are high due to limitations in handling overlapping ion chromatograms.

Conclusions:

• The ViMMS framework effectively simulates and compares DDA and DIA acquisition methods, with simulation results translating to real-world performance.
• Understanding the strengths and limitations of DDA and DIA based on ion complexity optimizes method selection for accurate metabolomics.
• ViMMS offers a resource-efficient platform for exploring and advancing LC-MS/MS data acquisition strategies.