Using XCiMS to Quantify the Benefits of Ion Mobility Spectrometry for Untargeted Metabolomics

Abstract

Integrating ion mobility spectrometry (IMS) with liquid chromatography-mass spectrometry (LC-MS) has the potential to improve untargeted metabolomics workflows, but assessing the benefits of IMS has been challenging due to a paucity of software solutions to process raw LC-IMS-MS data in a manner that enables direct comparison with LC-MS data. We developed the XCiMS software to collapse 4D data into pseudo-3D data. Peak detection is then performed in the pseudo-3D space, followed by targeted extraction of 4D signals from 3D-guided regions of interest. Here, we applied XCiMS to estimate the potential of IMS for untargeted metabolomics when analyzing human plasma. First, we assessed the number of metabolites that are uniquely resolved by LC-IMS-MS but not LC-MS. Although as many as 50% of 3D peaks produce multiple 4D peaks when using high-resolution IMS, few represent newly resolved structural isomers. We found that the increase in features due to IMS was mostly caused by charge isomers and ion clusters. Second, we investigated the impact of IMS on metabolite identification. When using a 16-minute chromatographic separation, we found that IMS can decrease the frequency of chimeric MS/MS by about half. We conclude that while IMS adds considerable complexity to untargeted metabolomics data without resolving a large number of additional structural isomers relative to a standard LC-MS experiment, it does facilitate metabolite identification by decreasing chimeric spectra and providing CCS values for library searching.