EVALUATION OF SEMI-TARGETED AND NON-TARGETED (‘DISCOVERY’) SCREENING TOOLS

Importance of Topic

The continuous emergence of novel psychoactive substances presents a significant challenge for toxicology laboratories. High-resolution mass spectrometry methods such as TOF-MSE provide comprehensive datasets that support both targeted screening and discovery-based investigations. Advanced informatics tools enable retrospective data mining and enhance the detection of known and unknown compounds.

Objectives and Study Overview

This study compares three screening workflows integrated in a single software platform: conventional targeted analysis, semi-targeted screening using in-silico fragmentation of Molfile structures, and a fully non-targeted Discovery approach that proposes elemental compositions, searches external libraries, and predicts fragment ions.

Methodology and Instrumentation

The evaluation used seventeen postmortem blood samples analyzed with an ACQUITY UPLC I-Class system coupled to a XEVO G2-XS QTOF mass spectrometer. Data were acquired in TOF-MSE mode under low and elevated collision energies to capture accurate precursor masses and diagnostic fragment ions. The UNIFI software performed automated targeted matching against a library of over 1300 entries, semi-targeted screening via Molfile-driven fragmentation, and non-targeted discovery workflows.

Key Results and Discussion

Targeted screening detected 138 instances of 74 toxicologically relevant substances, with 93% confirmed by at least one diagnostic fragment. Semi-targeted screening validated 88% of these detections and identified three additional compounds. The Discovery tool independently confirmed 85% of target identifications and provided structural proposals for potential unknowns.

Benefits and Practical Applications

The integrated workflows enhance screening confidence, especially for emerging substances lacking reference standards. Retrospective analysis enables continual reexamination of data as spectral libraries grow. Automation of multiple strategies streamlines high-throughput toxicology analysis.

Future Trends and Potential Applications

Ongoing developments may include expanded high-resolution spectral libraries enriched with retention time and fragment data, machine learning–driven prediction of fragmentation patterns, and broader applications to environmental monitoring, forensic toxicology, and clinical metabolomics.

Conclusion

Automated semi-targeted and non-targeted screening significantly improve the efficiency and reliability of mass spectrometry–based toxicology workflows, facilitating the detection and confirmation of both known and emerging psychoactive substances.

References

• European Monitoring Centre for Drugs and Drug Addiction (EMCDDA)
• United Nations Office on Drugs and Crime (UNODC)
• UK Home Office Forensic Early Warning System (FEWS)