Progenesis qi v2

The LC-MS raw data were analyzed by Progenesis QI v2.0 software (Waters Corporation, Milford, CT, USA) with a retention time (RT) range of 0.5–14 min, a mass range of 50–1000 Da, and a mass tolerance of 0.01 Da. Isotopic peaks were excluded from the analysis, the noise elimination level was set at 10, the minimum intensity was set to 15% of base peak intensity, and finally, the RT tolerance was set at 0.01 min. The LC-MS data were obtained with three-dimensional datasets included m/z, peak RT, and peak intensities, and RT-m/z pairs were used as the identifier for each ion.

The raw data collected by UPLC-Q-TOF/MS (ESI⁺ and ESI⁻ modes) into the Progenesis QI V2.0 software (Waters Corp., Milford, United States ) using baseline filtering, peak identification and correction functions to obtain a data matrix of retention time (t_R), mass-to-charge ratio and peak intensity. Multivariate statistical analysis was performed by EZinfo 3.0 software. First, unsupervised principal component analysis (PCA) was used to observe the overall distribution between groups of samples and the stability of the entire analysis process, and then supervised orthogonal partial least squares-discriminant analysis (OPLS-DA) was used to distinguish the overall differences in metabolic profiles between groups and to look for metabolites that differed between groups. The statistically significant differential metabolites (p < 0.05, fold changes ≥1.5, CV < 30%, VIP-value>1) and their ion fragments at high and low collision energy were matched with Human Metabolomics Database (HMDB) (https://hmdb.ca/) and self-built database to confirm the differential sphingolipid metabolites. Potential biomarkers associated with GE processing were analyzed by MetaboAnalyst 4.0 (http://www.metaboanalyst.ca) and Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.kegg.jp) and a metabolic network was constructed according to the relationships among the identified potential biomarkers.

Data obtained by QTOF were first converted into mzML format by ProteoWizard software and then processed in Progenesis QI v2.0 Software (Waters, Newcastle, UK). The optimal sample data was automatically selected by Progenesis QI to align all the remaining data. Subsequently, adduct ions were deconvoluted, and ion abundance above the threshold level was calculated. All features detected were matched and selected based on the Lipid Maps database within the mass tolerance of 10 ppm. The analyzed data was imported into Simca 14.1 (Umetrics, Umea, Sweden) for following multivariate statistical analysis such as orthogonal projections to latent structures-discriminate analysis (OPLS-DA) and principal component analysis (PCA). The ions with VIP values above 1.0 and p values of Student's t-test below 0.05 were selected as differential metabolites. These ions were further identified by tandem mass spectra.

The metabolomics data were preprocessed with Progenesis QI v2.3 software (Waters, Elstree, UK). The compound characterization was performed using the Human Metabolome Database (HMDB), LIPID MAPS (v2.3), and the METLIN database, and the metabolic pathway analyses were performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. The GraphPad Prism 8.0.1 (GraphPad Software, California, CA, USA), and R software packages (TUNA Team, Tsinghua University, Beijing, China) were used for the statistical analysis. The data are expressed as the mean ± standard deviation (SD). One−way ANOVA was performed to compare multiple groups.

MarkerLynx within MassLynx software version 4.1 (Waters Corp. Milford, United States) was employed to process all the collected UPLC-QTOF/MS data. Progenesis QI V 2.3 software (Waters, Milford, MA, United States) was used for peak detection and alignment, and EZinfo 3.0 software was used for the multivariate data analysis. To match and distinguish Rt-m/z pairs of different metabolites being selected in each group, an online database named HMDB (Human Metabolome Database, http://www.hmdb.ca/) was consulted and high-precision ion fragments were utilized. Enrichment analysis of the metabolomics pathways was performed through online websites, such as https://www.metaboanalyst.ca/and https://www.kegg.jp/, and the HMDB number of the identified differential metabolites were entered into the Metaboanalyst online data analysis software. Fisher’s exact test was selected as the enrichment method, and the rat Rattus norvegicus (rat) (KEGG) library was selected as the database for pathway analysis.
The data collected from the MWM and H&E staining were expressed as the mean ± standard deviation (SD), and the statistical significance was measured by GraphPad Prism 8.0.1. Spearman’s correlation analysis of the GM and endogenous metabolites was performed by the OmicStudio tools at https://www.omicstudio.cn/tool.