LipidMS 3.0: an R-package and a web-based tool for LC-MS/MS data processing and lipid annotation.

Abstract

LipidMS was initially envisioned to use fragmentation rules and data-independent acquisition (DIA) for lipid annotation. However, data-dependent acquisition (DDA) remains the most widespread acquisition mode for untargeted LC-MS/MS-based lipidomics. Here, we present LipidMS 3.0, an R package that not only adds DDA and new lipid classes to its pipeline but also the required functionalities to cover the whole data analysis workflow from pre-processing (i.e. peak-peaking, alignment and grouping) to lipid annotation. We applied the new workflow in the data analysis of a commercial human serum pool spiked with 68 representative lipid standards acquired in full scan, DDA and DIA modes. When focusing on the detected lipid standard features and total identified lipids, LipidMS 3.0 data pre-processing performance is similar to XCMS, whereas it complements the annotations returned by MS-DIAL, providing a higher level of structural information and a lower number of incorrect annotations. To extend and facilitate LipidMS 3.0 usage among less experienced R-programming users, the workflow is also implemented as a web-based application. The LipidMS R-package is freely available at https://CRAN.R-project.org/package=LipidMS and as a website at http://www.lipidms.com. Supplementary data are available at Bioinformatics online.