Identifying Food and Environmental Contaminants using the New NIST High-Res MS/MS Library Search Algorithms and Publicly Available LC/MS/MS Spectral Libraries

Importance of the Topic

High-resolution tandem mass spectral libraries are crucial for reliable identification of trace contaminants in food and environmental samples.

Objectives and Study Overview

This study evaluates the performance of NIST high-resolution MS/MS library search algorithms using publicly available LC/MS/MS spectral libraries to identify pesticides, veterinary drugs and personal care products in spiked samples.

Methodology and Instrumentation

Approximately 100 000 spectra were retrieved from MoNA and other public repositories and converted into NIST user libraries. Known contaminants were spiked into solvent and analyzed on Agilent Q-TOF LC/MS in Auto MS/MS mode. Data processing used Agilent MassHunter Qualitative Analysis and NIST MS Search v2.3. Identification confidence was assessed by top hit rankings and ROC curve analysis of DotProd and Rev-Dot scores.

Instrumentation Used

• Agilent Q-TOF LC/MS instruments with Auto MS/MS acquisition
• Agilent MassHunter Qualitative Analysis Software v10.0
• NIST MS Search Program v2.3
• Public spectral repositories: MoNA, MassBank EU, Vaniya/Fiehn Natural Products Library, ReSpect, HMDB, MetaboBASE, GNPS

Key Results and Discussion

Rev-Dot achieved 93 % correct top 1 identifications and 100 % in top 3, outperforming DotProd which showed only ~50 % accuracy. ROC analysis yielded an AUC of 0.90 for Rev-Dot versus 0.33 for DotProd. A 10 ppm precursor mass tolerance provided mostly unique matches. The HiRes hybrid search successfully proposed structurally related analogues for compounds absent from the library.

Benefits and Practical Applications

NIST MS/MS library search with public HRAM databases enables robust identification of food and environmental contaminants. The hybrid similarity search expands capabilities for unknown identification by highlighting substructure relationships.

Future Trends and Potential Applications

Continuous growth of crowd-sourced high-resolution spectral libraries will enhance suspect and non-target screening workflows. Advances in scoring algorithms and hybrid search strategies are expected to further improve compound identification in complex matrices.

Conclusion

NIST high-resolution MS/MS search algorithms applied to crowd-sourced libraries deliver high-confidence identifications of contaminants. Rev-Dot scoring provides a reliable metric, and the hybrid search extends coverage to novel or unrepresented compounds.

Reference

• Stein SE. NIST Mass Spectral Search Program v2.3 User’s Guide
• MassBank of North America (MoNA) repository
• Chao A et al., Anal Bioanal Chem 412, 1303 (2020)
• MassBank EU, ReSpect, HMDB, MetaboBASE, GNPS public spectral libraries