Novel data processing workflow including in-silico derivatization for integrated CI & EI MS data acquired with the GC-ecTOFinstrument

Significance of the topic

Derivatized GC-MS workflows that combine Chemical Ionization (CI) and Electron Ionization (EI) are increasingly important for non-target and suspect screening of volatile and semi-volatile compounds in fields such as environmental monitoring, food and flavor analysis, and metabolomics. The simultaneous acquisition of CI and EI data on a single platform increases identification confidence by providing complementary information: intact (or quasi-molecular) ions and isotopic patterns from CI together with structural fragment information from EI. A robust data-processing pipeline that integrates both ionization modes and supports in-silico derivatization and fragmentation expands the capability to annotate compounds even when reference standards or nominal-mass EI library matches are unavailable.

Objectives and study overview

The work presents a novel processing workflow for GC-ecTOF data that: identifies features in CI data in an untargeted manner; reconstructs coeluting EI fragmentation spectra for those features; applies automated in-silico derivatization and in-silico fragmentation for candidate annotation; and transfers confident annotations into targeted screening and quantitation methods. The pipeline is demonstrated on two application studies using derivatized biological matrices (mouse wound extracts and human fecal reference material) to illustrate profiling, annotation, and the transition to routine TASQ-based screening and quantification.

Methods and workflow

The workflow consists of sequential steps implemented in prototype versions of MetaboScape and TASQ after conversion of raw GC-ecTOF data to TSF format. Key processing steps are:

• Untargeted feature detection performed on the CI data stream with de-isotoping and assembly of CI-specific adduct/fragment ion groups (typical ions considered: [M-CH3]+, [M-H]⁺, [M]⁺, [M+H]⁺, [M+NH4]⁺).
• Cross-sample registration and retention time alignment across the study set to produce a consistent feature table.
• Reconstruction of EI fragmentation spectra for CI-detected features by computing a weighted-average EI spectrum from EI scans that coelute with the CI monoisotopic extracted ion chromatogram (EIC).
• Automated annotation using a target list (name, neutral formula, InChI/SMILES) that is expanded on-the-fly by in-silico derivatization to generate expected derivatized species for each compound.
• Application of in-silico fragmentation to the derivatized candidate structures to produce predicted EI fragments for multi-criteria matching (accurate mass, isotopic pattern, and fragment agreement).
• Annotation acceptance thresholds reported include narrow and wide mass accuracy windows (2.0 and 5.0 mDa) and mSigma thresholds for isotopic pattern matching (20 and 100), allowing flexible stringency levels.

Instrumentation used

• GC-ecTOF platform (Bruker Daltonics) enabling simultaneous CI and EI acquisition.
• Data conversion to TSF format prior to processing.
• MetaboScape (prototype) used for untargeted CI feature finding, retention time alignment, in-silico derivatization, in-silico fragmentation, and reconstruction/assignment of EI spectra to CI features.
• TASQ software (prototype) used to generate target screening/quantitation methods from MetaboScape annotations and to automate batch acquisition and processing.
• Derivatization chemistries used in studies: two-step methoximation (MeOX) followed by MSTFA for mouse wound samples; MTBSTFA for fecal samples. NIST SRM 8048 human fecal reference material informed feces compound lists.

Main results and discussion

• Untargeted feature detection on CI data provided high-quality monoisotopic masses and isotopic patterns, enabling robust cross-sample feature registration and retention time alignment.
• Reconstruction of EI spectra tied to CI features produced weighted-average EI fragmentation spectra referenced to the CI monoisotopic EIC, facilitating assignment of EI fragments to the same chromatographic signal as the CI-derived molecular ion.
• On-the-fly in-silico derivatization expanded the searchable target list to include expected derivatized products, addressing a major limitation of nominal-mass EI spectral libraries for derivatized samples.
• Combining accurate-mass CI ions, isotopic fidelity, and reconstructed EI fragments enabled multi-criteria annotations that raised confidence for compounds lacking authentic standards.
• Automated export of annotations from MetaboScape into TASQ generated ready-to-run target screening and quantitation methods (including name, formula, structure, retention time, accurate mass definitions for CI ions, and selected abundant EI fragments), smoothing the path from discovery to routine assays.

Benefits and practical applications

• The integrated CI+EI workflow increases identification confidence in non-targeted GC-MS profiling by providing orthogonal evidence: intact/quasi-molecular ion information from CI and diagnostic fragments from EI.
• In-silico derivatization allows annotation of derivatized analytes without prior library spectra, expanding coverage for metabolomics, environmental monitoring, food/flavor analysis, and forensic or clinical research contexts (research use only).
• Automated transfer to TASQ supports scalable screening workflows and routine quantitation, enabling laboratories to convert discovery markers into validated screening methods more rapidly.

Future trends and applications

• Improved in-silico chemistry models: expanding derivatization rules and reaction-specific fragment prediction (including isotope- and adduct-specific behavior) will strengthen annotation reliability for more diverse chemistries.
• Machine-learning-assisted scoring integrating CI mass accuracy, isotopic fit, reconstructed EI fragment patterns, and chromatographic behavior could further prioritize correct annotations and reduce false positives.
• Community-curated derivatized EI spectral resources and harmonized data formats would accelerate method transfer and cross-laboratory reproducibility.
• Integration with suspect lists from reference materials and databases (e.g., SRMs) combined with automated reporting will enhance applicability in regulatory and routine monitoring programs.

Conclusion

The presented GC-ecTOF processing pipeline demonstrates a practical route from untargeted CI-based feature discovery to confident compound annotation by leveraging reconstructed EI spectra and automated in-silico derivatization/fragmentation. The seamless handoff of annotations into TASQ for targeted screening and quantitation supports rapid translation of discovery results into routine assays. Together, these capabilities extend the reach of GC-MS metabolomics and non-target screening for derivatized samples, particularly when reference standards are not available. Users should note the research-use status of the software and the authors' affiliations with the instrument vendor.

References

1. NIST Reference Material 8048 Human Fecal Material — Reference Material Information Sheet. NIST SRM 8048.
2. Ruttkies C, Schymanski EL, Wolf S, Hollender J, Neumann S (2016) MetFrag relaunched: incorporating strategies beyond in silico fragmentation. Journal of Cheminformatics 8:3. DOI: 10.1186/s13321-016-0115-9.