Automating component detection of small molecules in complex mixtures using HRAM Q-TOF data

Importance of the Topic

High‐resolution accurate mass quadrupole time‐of‐flight (HRAM Q‐TOF) mass spectrometry is increasingly essential for rapid and comprehensive detection of small molecules in complex matrices. Automating component detection streamlines data processing, reduces manual interpretation time and improves consistency, making suspect screening more robust in fields such as food safety and forensic toxicology.

Objectives and Study Overview

This work presents a novel algorithm designed to automatically locate and characterize chromatographic components in HRAM Q‐TOF full scan data. The algorithm was evaluated on two complex sample types: QuEChERS extracts of food commodities spiked with a panel of pesticides, and human plasma spiked with drugs of abuse (DoA). Performance was assessed by matching detected features against targeted compound lists.

Methodology and Instrumentation

A generic LC–MS/MS method was employed: a Nexera LC system coupled to an LCMS-9030 QTOF (Shimadzu). Two columns were used—Restek Raptor ARC18 for pesticides and Restek Biphenyl for DoA screening. Data acquisition combined a full MS scan (m/z 140–900 or 100–500) with sequential data‐independent DIA‐MS/MS scans using 20 Da windows. Chromatographic peak width (FWHM) was the primary parameter for component detection alongside minimal intensity thresholds.

Main Results and Discussion

• Food Safety Screening: From an apple extract spiked at 0.1 mg/kg, the algorithm grouped 3 100 features (including adducts, neutral losses and charge variants). A target list of 212 pesticides returned 207 correct matches within 5 ppm mass error and 0.25 min retention time tolerance.
• Toxicology Screening: From plasma spiked at 50 ng/mL, 6 223 grouped features were detected. A filtered library of 153 DoA compounds delivered 45 confirmed hits, all within 5 ppm (or <1 mDa below m/z 200) and 0.2 min retention time window.
• Algorithm Workflow: A two‐step approach (‘search’ and ‘validation’) first correlates masses across a moving FWHM window to propose chromatographic peaks, then refines isotopic patterns and merges related ion species into single component outputs, reducing false positives.

Benefits and Practical Applications

• Dramatically reduced manual review time through automated peak picking and spectral interpretation.
• Simplified reporting of monoisotopic masses by consolidating multiple ionization species.
• Flexibility for both untargeted monitoring and targeted suspect screening in diverse complex matrices.
• Improved confidence in compound identification via integrated adduct and isotopic validation.

Future Trends and Applications

Advances may include integration with machine learning models for predictive library search, real‐time data processing during acquisition, and expanded workflows for metabolomics or environmental analysis. Enhanced spectral libraries and dynamic tolerance settings could further improve the sensitivity and selectivity of automated screening.

Conclusion

The presented component detection algorithm effectively automates the identification of small molecules in complex HRAM Q‐TOF data. Its robust performance in food safety and forensic toxicology demonstrates its versatility for both targeted and untargeted workflows, offering a scalable solution for high‐throughput screening.

References

Simon Ashton, Kirsten Hobby, Alan Barnes, Neil Loftus. Automating component detection of small molecules in complex mixtures using HRAM Q-TOF data. ASMS 2019, Shimadzu Corporation.