Implementing Annotation Confidence Scoring in Untargeted Mass Spectrometry Workflows for Small Molecule Analysis.

Area of Science:

• Environmental science
• Biological research
• Metabolomics

Background:

• Untargeted small molecule analysis using high-resolution mass spectrometry is crucial for complex sample characterization.
• Interpreting and reporting data from these analyses is challenging due to high dimensionality and lack of standardized annotation confidence frameworks.
• Existing platforms lack automated systems for ranking annotation confidence, hindering cross-study comparability.

Purpose of the Study:

• To develop and validate a novel method for assigning and reporting annotation confidence levels in untargeted small molecule analysis.
• To enhance the specificity and transparency of data reporting in metabolomics and related fields.
• To create a standardized framework for evaluating the reliability of compound identifications.

Main Methods:

• A custom scripting node was developed, integrating into commercial data analysis software as a postprocessing tool.
• The script assigns annotation confidence based on the Schymanski et al. scoring scheme, expanding it with four new sublevels.
• Annotation confidence is assessed using consensus scoring across multiple compound identification workflow nodes (e.g., mzCloud, ChemSpider).

Main Results:

• The developed script demonstrated high scoring accuracy: 100% in negative mode and >99.5% in positive mode for NIST SRM 1950 plasma samples.
• Validation was performed across both Reversed-Phase Liquid Chromatography (RPLC) and Hydrophilic Interaction Liquid Chromatography (HILIC) separations.
• The tool successfully improved specificity by distinguishing between established annotation confidence categories.

Conclusions:

• The developed scripting node provides a standardized and automated approach to assigning annotation confidence in untargeted small molecule analysis.
• This tool enhances data transparency, promotes consistency across studies, and facilitates the standardization and comparability of metabolomics results.
• The scoring approach is broadly applicable beyond metabolomics to other untargeted small molecule workflows.