Enhancing untargeted lipidomics using offline two-dimensional LC-MS and systematic evaluation of data processing tools

Although two-dimensional liquid chromatography-mass spectrometry (2DLC-MS) has been explored to improve lipidomic coverage, its practical implementation remains limited by challenges in fractionation design and interpretation of complex datasets. Here we report an optimized offline hydrophilic interaction chromatography-reversed-phase liquid chromatography (HILIC-RPLC) 2DLC-MS workflow for untargeted lipidomics. A simplified five-fraction HILIC strategy was developed to enhance lipid class separation while maintaining practical analytical throughput. Optimization of solvent composition, mobile phases, and gradient conditions improved compatibility between chromatographic dimensions and enabled efficient fractionation prior to RPLC-MS analysis. Compared with conventional one-dimensional RPLC-MS, the proposed workflow reduced lipid co-elution and ion suppression, resulting in approximately two-fold higher lipid feature coverage in biological samples including serum, tissue, and cultured cells. Chromatographic fractionation information was further used as an orthogonal parameter to evaluate lipid annotation reliability. Using this framework, three lipidomics software platforms-MetaboScape, MS-DIAL, and LipidScreener-were systematically evaluated for processing 2DLC-MS datasets. The results reveal substantial variability among software tools and demonstrate that integrating chromatographic fractionation with data processing improves confidence in untargeted lipidomic analysis.