A Discovery Lipidomic Workflow Combining the Xevo™ MRT P10 with mzmine Data Processing Pipelines

Importance of the Topic

Lipidomics requires analytical approaches that balance broad chemical coverage with confident structural annotation. Robust acquisition strategies and transparent data processing are critical for detecting low-abundance species, distinguishing isobaric/isomeric lipids, and ensuring reproducible identifications in biological and clinical studies. Integrating high-resolution MS instrumentation with advanced, open-source data pipelines addresses these needs and enables reliable exploration of complex lipidomes.

Objectives and Study Overview

This study evaluated a discovery lipidomics workflow combining the Xevo MRT P10 multi-reflecting TOF mass spectrometer with an mzmine-based processing pipeline. The goals were to compare data-dependent acquisition (DDA) and data-independent acquisition (DIA) on the MRT platform, assess identification coverage and annotation confidence, and demonstrate how a suite of quality-control and annotation tools (rule-based scoring, ECN modelling, Kendrick Mass Defect analysis, and Interactive Molecular Networking) improves lipidome characterization. Human plasma standard NIST SRM 1950 was used as the test material.

Methodology

• Sample preparation: Lipids extracted from NIST SRM 1950 plasma using the Bligh–Dyer method. Extracts were reconstituted in methanol and injected in technical triplicate (1 µL for positive ionization, 2 µL for negative).
• Chromatography: Reversed-phase UPLC using a short 12-minute cycle on an ACQUITY Premier CSH C18 column (1.7 µm, 2.1 × 100 mm). Gradient from 50% to 99% organic (IPA/ACN/10 mM ammonium formate with 0.1% formic acid) with rapid re-equilibration to enable high-throughput runs.
• Mass spectrometry — DDA: MS1 at 10 Hz and MS/MS at 20 Hz; top‑30 selection for positive mode, top‑20 for negative; dynamic exclusion tolerance ~2 s and 5 ppm; quadrupole isolation window 1 Da. Source/desolvation parameters were optimized for MRT operation (capillary voltages ~2.5/1.2 kV for ESI+/ESI-, cone 30 V, desolvation 500 °C, desolvation gas 700 L/min).
• Mass spectrometry — DIA (MSE): Alternating low/high energy acquisition with 20 Hz aggregate scan rate (10 Hz MS1 + 10 Hz MS2), collision energy ramp 25–55 eV, mass range 50–1200 m/z.
• Data export and preprocessing: Raw data automatically exported to mzML via waters_connect. mzmine pipeline included peak detection, smoothing, isotope filtering, and (for DIA) spectral deconvolution and pseudo‑MS2 reconstruction. Postprocessing steps included gap-filling, duplicate feature filtering, and Ion Identity Networking to group adducts and related in-source species.

Used Instrumentation

• XEVO MRT P10 Multi-Reflecting Time-of-Flight Mass Spectrometer (high resolution, high mass accuracy MS and MS/MS).
• ACQUITY Premier UPLC system equipped with ACQUITY Premier CSH C18 column (1.7 µm, 2.1 × 100 mm).
• waters_connect software (automated mzML export) and mzmine (open-source processing, annotation, and QC modules).

Data Processing and Annotation Strategy

• Rule-based lipid annotation combined with multi-criteria scoring to increase confidence in identifications.
• ECN (Equivalent Carbon Number) retention time modelling used to verify class-specific chromatographic behavior and detect outliers or misassignments.
• Kendrick Mass Defect (KMD) analysis to visualize homologous series and support class grouping.
• Interactive Molecular Networking based on MS/MS similarity to cluster lipids by class and propagate annotations to otherwise unannotated features.
• Lipid Annotation QC Dashboard to review RT distributions, isotope patterns, diagnostic fragments, duplicate annotations, and overall annotation quality.

Main Results and Discussion

• Both acquisition modes produced high-quality, reproducible MS/MS data suitable for reliable rule-based lipid annotation when processed through the mzmine workflow.
• DDA identified a greater total number of lipid species and required shorter processing times compared to DIA, reflecting the higher spectral specificity of DDA for prioritized precursors.
• DIA provided improved coverage for certain sphingolipid subclasses (notably hexosylceramides) likely due to its systematic fragmentation across the entire mass range and better sampling of low-abundance precursors.
• ECN and KMD plots effectively highlighted homologous series and verified retention behavior across subclasses; Interactive Molecular Networking revealed clusters of chemically related species and helped detect missing or inconsistent annotations.
• Overall fragmentation quality between DDA and DIA was comparable for annotation purposes, but their complementary strengths (DDA: depth and speed; DIA: comprehensive coverage of low-abundance or stochastically missed species) recommend combined consideration depending on study goals.

Benefits and Practical Applications

• The integrated MRT + mzmine workflow facilitates transparent, reproducible lipidomics with built-in QC and visualization tools, suitable for discovery studies and downstream targeted follow-up.
• Rapid LC turnaround and automated mzML export support higher throughput laboratories and standardized data pipelines.
• ECN, KMD and spectral networking reduce false positives and improve annotation propagation, aiding biomarker discovery, clinical lipid profiling, and mechanistic lipid biology studies.
• DDA is advantageous for routine identification-rich experiments and faster processing; DIA is recommended when exhaustive coverage of low-abundance subclasses (e.g., specific sphingolipids) is required.

Future Trends and Opportunities

• Further integration of open-source processing with vendor-ready export (automated mzML) will promote reproducibility and data sharing across laboratories.
• Improvements in DIA deconvolution algorithms and hybrid acquisition schemes could close the gap in identification counts while preserving DIA’s systematic coverage.
• Machine learning–driven annotation scoring and automated isomer discrimination (e.g., ion mobility, retention modelling combined with MS/MS patterns) will enhance structural resolution and confidence.
• Standardized QC dashboards and community-curated lipid libraries will accelerate cross-study comparisons and regulatory acceptance of lipidomic biomarkers.

Conclusion

Combining the Xevo MRT P10 mass spectrometer with an mzmine-based data-processing pipeline yields a robust discovery lipidomics workflow. DDA offers higher identification counts and faster processing, while DIA complements it by increasing coverage for certain low-abundance sphingolipids. The use of rule-based annotation, ECN modelling, KMD analysis, Interactive Molecular Networking, and a dedicated QC dashboard produces reproducible, high-confidence lipid annotations suitable for discovery and comparative lipidomics.

Reference

• Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Canadian Journal of Biochemistry and Physiology. 1959;37(8):911–917.