LAUSR

Lipidomics is a subfield of metabolic phenotyping that focuses on high-throughput profiling and quantification of lipids. Essential roles of lipidomics in translational and clinical research have emerged, especially over the past decade. Most lipidomic pipelines have been developed using mass spectrometry (MS)-based methods. Because of the complexity of the data, generally, computational demands are much higher in untargeted lipidomic studies. In the current paper, we primarily discussed the recent advances in untargeted liquid chromatography–mass spectrometry-based lipidomics, covering various facets from analytical strategies to functional interpretations. The current practice of tandem MS-based lipid annotation in untargeted lipidomics studies was demonstrated. Notably, we highlighted the essential characteristics of machine learning models, together with a data partitioning strategy, to facilitate appropriate modeling and validation in metabolic phenotyping studies. Critical aspects of data sharing were briefly mentioned. Finally, certain recommendations were suggested toward more standardized and sustainable lipidomics analysis strategies as independent platforms, and as members of the omics family.