Batchwise data analysis with inter-batch feature alignment in large scale platelet lipidomics study using...

Area of Science:

• Biochemistry
• Analytical Chemistry
• Clinical Research

Background:

• Untargeted lipidomics via UHPLC-MS/MS with data-independent acquisition (DIA) is valuable for clinical research.
• Large-scale studies (1000s of samples) face challenges in data processing due to batch effects like retention time and mass shifts.
• Limited computer power exacerbates difficulties in processing extensive datasets over time.

Purpose of the Study:

• To develop and validate a batchwise data processing strategy for untargeted lipidomics using DIA.
• To improve lipidome coverage and feature identification in large clinical cohorts.
• To establish a robust workflow for handling extensive lipidomics data from multiple batches.

Main Methods:

• Implemented a batchwise data processing strategy using MS-DIAL for automated analysis.
• Utilized inter-batch feature alignment based on precursor m/z and retention time similarity.
• Established a reference peak list for targeted data extraction from combined batches.

Main Results:

• The workflow was validated on platelet lipid extracts from coronary artery disease (CAD) patients.
• Applied to a cohort of 1057 CAD patients across 22 batches, significantly increasing lipidome coverage.
• Optimal increase in annotated features was observed using 7-8 batches, improving structural identification.

Conclusions:

• Batchwise processing with inter-batch feature alignment is effective for large-scale untargeted lipidomics.
• This strategy enhances lipidome coverage and identification accuracy in clinical studies.
• The method provides a robust solution for managing and analyzing extensive DIA-MS/MS datasets.