Artificial Intelligence-Based LC-OzESI-MRM for Isomer-Resolved Triacylglycerol Profiling by In-Source Ozonolysis

The structural complexity of triacylglycerols (TGs) and the positional specificity of carbon-carbon double bonds (C═C) within fatty acyl (FA) chains present a formidable challenge in structural lipidomics. Traditional tandem mass spectrometry (MS) approaches often lack the capability to distinguish TG isomers, limiting insights into lipid biochemistry and isomer-specific functionality. Here, we introduce an online liquid chromatography (LC) approach that enables isomer-specific TG identification through in-source ozonolysis via ozone electrospray ionization (OzESI) combined with multiple-reaction monitoring (MRM) on a triple quadrupole mass spectrometer. By directly incorporating ozone into the electrospray ionization source, our approach generates diagnostic ozonolysis product ions that localize C═C positions without the need for post-ionization reaction chambers or major instrument modifications. We demonstrate the utility of this method using synthetic TG standards and apply it to real-world samples, including the analysis of canola oil across processing stages. To overcome the challenges of data analytics, automated data processing is achieved via CLAW-OzESI-MRM (Comprehensive Lipidomics Automated Workflow for OzESI-MRMs) which includes an artificially intelligent (AI) agentic framework to streamline structural assignment of TGs with C═C specificity. This robust, accessible platform offers enhanced isomeric resolution for lipidomics research, expanding the analytical capabilities of targeted MS workflows in food, biomedical, and industrial applications.