Customized Consensus Spectral Library Building for Untargeted Quantitative Metabolomics Analysis with Data Independent Acquisition Mass Spectrometry and MetaboDIA Workflow

Data independent acquisition-mass spectrometry (DIA-MS) coupled with liquid chromatography is a promising approach for rapid, automatic sampling of MS/MS data in untargeted metabolomics. However, wide isolation windows in DIA-MS generate MS/MS spectra containing a mixed population of fragment ions together with their precursor ions. This precursor-fragment ion map in a comprehensive MS/MS spectral library is crucial for relative quantification of fragment ions uniquely representative of each precursor ion. However, existing reference libraries are not sufficient for this purpose since the fragmentation patterns of small molecules can vary in different instrument setups. Here we developed a bioinformatics workflow called MetaboDIA to build customized MS/MS spectral libraries using a user's own data dependent acquisition (DDA) data and to perform MS/MS-based quantification with DIA data, thus complementing conventional MS1-based quantification. MetaboDIA also allows users to build a spectral library directly from DIA data in studies of a large sample size. Using a marine algae data set, we show that quantification of fragment ions extracted with a customized MS/MS library can provide as reliable quantitative data as the direct quantification of precursor ions based on MS1 data. To test its applicability in complex samples, we applied MetaboDIA to a clinical serum metabolomics data set, where we built a DDA-based spectral library containing consensus spectra for 1829 compounds. We performed fragment ion quantification using DIA data using this library, yielding sensitive differential expression analysis.