MSⁿLib: efficient generation of open multi-stage fragmentation mass spectral libraries

Area of Science:

• Analytical Chemistry
• Biochemistry
• Computational Chemistry

Background:

• Untargeted high-resolution mass spectrometry is crucial for metabolomics, natural product discovery, and exposomics.
• Compound identification is a significant challenge, with current methods relying on spectral library matching of tandem mass spectrometry (MS²) data.

Purpose of the Study:

• To address the lack of open multi-stage fragmentation (MSⁿ) reference data for diverse chemicals.
• To develop a machine learning-ready resource to improve compound identification.

Main Methods:

• Creation of MSⁿLib, an open resource containing over 2 million spectra.
• Utilized a high-throughput data acquisition and processing pipeline.
• Employed the open-source software mzmine for data processing.

Main Results:

• MSⁿLib comprises MSⁿ trees for 30,008 unique small molecules.
• The resource is designed to be machine learning-ready, facilitating advanced data analysis.
• Provides deeper insights into substructures and enables validation of fragmentation pathways.

Conclusions:

• MSⁿLib significantly enhances the available reference data for MSⁿ analysis.
• This resource is expected to accelerate compound identification in various fields, including metabolomics and natural product discovery.
• Facilitates the validation of fragmentation pathways and substructure analysis.