A novel algorithm for automating fragment ion structure assignment using high mass accuracy MS/MS data

Importance of the topic

A clear understanding of fragment ion structures in high-mass accuracy MS/MS data is essential for reliable compound identification and verification across food safety, toxicology and metabolomics. Automating this annotation accelerates data interpretation, reduces manual errors and supports rapid decision making in QA/QC and research environments.

Objectives and Study Overview

The primary goal was to develop and test an algorithm that automatically assigns chemical structures to fragment ions observed in accurate mass Q-TOF MS/MS spectra. Key aspects:

• Demonstrate automated annotation on chemically diverse small molecules.
• Validate the algorithm in complex matrices: food extracts, human plasma and lipidomics samples.
• Evaluate performance under data-independent acquisition (DIA-MS/MS) with broad collision energy ranges.

Methodology and Used Instrumentation

The algorithm workflow:

• Generate all ionized tautomeric forms of the precursor.
• Identify atom groupings susceptible to bond cleavage or rearrangement.
• Simulate homolytic/heterolytic cleavages, ring openings and intramolecular shifts with neutral losses (NH₃, CO, HCN).
• Filter hypotheses by chemical charge and valence rules.
• Combine multistage fragmentation pathways to predict observed fragments.

Used instrumentation:

• Shimadzu LCMS-9030 Q-TOF system operating in MS and DIA-MS/MS modes.
• Cycle times under 1 s, mass range 40–1 000 Da, isolation width 20 m/z, collision energy spread 0–30 V.
• External mass calibration for sub-ppm mass accuracy.

Key Results and Discussion

The algorithm was applied to three case studies:

• Food safety (ethiofencarb vs. methiocarb): Differentiation via key fragments at m/z 164 (ethiofencarb) and m/z 121 (methiocarb) explained by carbamate loss and rearrangement.
• Drugs of abuse screening (cocaine vs. fentanyl in plasma): Automated assignment of characteristic product ions (m/z 182, 150 for cocaine; m/z 188, 216 for fentanyl) at 5 ng/mL enabled confident isomer discrimination.
• Metabolomics (lyso-phosphatidylcholine 16:0): Annotation of headgroup ions at m/z 104, 124 and 184 confirmed sn-position and fatty acyl composition.

This approach captured diverse fragmentation mechanisms including single-bond cleavages, rearrangements and neutral losses, achieving annotation consistent with manual interpretation.

Benefits and Practical Applications

Key advantages of the automated fragment annotation algorithm:

• High throughput: Annotates hundreds of spectra in minutes.
• Consistency: Reduces variability associated with manual assignments.
• Versatility: Applicable to pesticides, pharmaceuticals and lipid species in complex matrices.
• Enhanced confidence: Supports regulatory and research workflows by verifying compound identity with structural evidence.

Future Trends and Applications

Advances may include integration with machine-learning classifiers to refine fragmentation models, expansion to larger biomolecules (peptides, glycomics), and real-time assignment directly on instrument software. Coupling with spectral libraries and network-based visualization will further enhance interpretability and discovery in untargeted analyses.

Conclusion

The novel algorithm successfully automates structural assignment of high-accuracy MS/MS fragment ions across diverse compound classes. Its implementation on the LCMS-9030 Q-TOF platform streamlines method development and strengthens compound verification in food safety, toxicology and metabolomics applications.

Reference

• Nunez et al., Toxics, 2018.