An actionable annotation scoring framework for gas chromatography-high-resolution mass spectrometry

Area of Science:

• Environmental health sciences
• Analytical chemistry
• Toxicology
• Omics technologies

Background:

• Omics technologies, including exposomics, enable comprehensive characterization of environmental chemical exposures and their molecular biological responses.
• Standardization and confidence assignment in chemical annotations are critical limitations for advancing metabolomics and exposomics research.
• Gas chromatography high-resolution mass spectrometry (GC-HRMS) generates extensive chemical data, but its incorporation into annotation frameworks and confidence communication remains challenging.

Purpose of the Study:

• To develop and validate a qualitative framework for assigning confidence levels to chemical annotations derived from GC-HRMS data.
• To adapt and extend existing scoring schemas (Schymanski) for reporting confidence in GC-HRMS data analysis.
• To facilitate the comparison of chemical data across different platforms and studies in exposomics research.

Main Methods:

• The study discusses key evidence from common GC-HRMS workflows, including retention time, spectral matching (EI, PCI, ECNI, APCI), molecular ion, accurate mass, isotopic patterns, database occurrence, and blank occurrence.
• A qualitative framework is proposed, adapting the Schymanski scoring schema for GC-HRMS data to communicate confidence in chemical annotations.
• The framework was validated using spiked plasma samples and analysis of outdoor and indoor air samples, assessing false-positive rates for suspect screening.

Main Results:

• The proposed framework incorporates multiple lines of evidence to assign confidence levels to GC-HRMS chemical identifications.
• Validation demonstrated a 12% false-positive rate for Level 2 identifications in suspect screening (excluding structurally similar isomers).
• The framework is adaptable to various workflows and provides a concise method for communicating annotation confidence.

Conclusions:

• The developed framework offers a standardized approach to communicate confidence in GC-HRMS annotations, crucial for exposomics research.
• Adoption and further validation of this framework can lead to harmonization across the field, enhancing data quality and interpretability.
• Improved confidence in chemical annotations will advance the understanding of chemical exposomes and their impact on human health.