Untargeted liquid chromatography coupled to mass spectrometry (LC-MS) analysis generates massive amounts of information-rich mass data which presents storage and processing challenges. In this work, the validation of a recently proposed procedure for LC-MS data compression and processing is presented, using as example the analysis of lipid mixtures. This method consists of a preliminary selection of the Regions of Interest of the LC-MS data (MSROI) coupled to their throughout chemometric analysis by the Multivariate Curve Resolution Alternating Least Squares method (MCR-ALS). The proposed data selection procedure is based on the search of the most significant mass traces regions with high mass densities. This allows for a drastic reduction of the MS data size and of the computer storage requirements, without any significant loss neither of spectral resolution nor of accuracy on m/z measures. The combination of the MSROI data compression and MCR-ALS data analysis procedures in the new ROIMCR procedure has the main advantage of not requiring neither the chromatographic peak alignment nor the chromatographic peak shape modelling used in many other procedures as a pre-treatment step of the data analysis. The proposed ROIMCR procedure is tested in the analysis of the LC-MS experimental data coming from different lipid mixtures and of a melanoma cell line culture sample with satisfactory results. The proposed strategy is shown to be a general, fast, reliable and easy to use method for general untargeted LC-MS metabolic and lipidomic data analysis type of studies.
Keywords: Lipidomic analysis; ROIMCR; Validation.
Copyright © 2018. Published by Elsevier B.V.