Ultrahigh-performance liquid chromatography equipped with quadrupole time of flight mass spectrometry (UHPLC-Q-TOF/MS) analysis

Serum samples were separated by the ultra-high-performance liquid chromatography (UHPLC) system (Agilent 1290, Agilent Technologies, Palo Alto, USA) incorporating a hydrophilic interaction liquid chromatography (HILIC) column (2.1 mm × 100 mm, 1.7 μm; Waters, Milford, MA, USA). The samples were analyzed using a triple time-of-flight (TOF) mass spectrometer (ESI/Triple TOF 5600; AB Sciex, Concord, Canada) equipped with an electrospray ionization source used in positive and negative ion modes. The pretreatment, extraction, and identification of serum samples were according to the procedure described by Hu et al. [27]. The raw data (whiff scan files) were converted into mzXML format using ProteoWizard [28] and were imported to the XCMS software for peak matching, retention time alignment, and peak area extraction [29]. Metabolite structure identification was performed by comparing the accuracy of m/z values (< 25 ppm), and MS/MS spectra were interpreted with an in-house database (Shanghai Applied Protein Technology Co. Ltd., China) established with authentic standards. For the XCMS data, the ion peaks that were missing values greater than 50% in the group were filtered and excluded and data were normalized to total peak intensity. Then, the pattern recognition was analyzed by SIMCA-P software (version 14.1, Umetrics, Umea, Sweden), where could performed to multivariate data measurement, including unsupervised principal component analysis (PCA), supervised partial least squares discriminant analysis (PLS-DA), and supervised orthogonal partial least squares discriminant analysis (OPLS-DA), which were carried out to uncover and extract the statistically significant metabolite variations. The PLS-DA and OPLS-DA models were validated based on multiple correlation coefficient (R²) and forecast ability according to the model (Q²) in cross-validation and permutation test by applying 200 iterations [30]. The R² value in the permutated plot described how well the data fit the derived model, whereas the Q² value described the predictive ability of the constructed model and was a measure of model quality [31]. Volcano Plot measurement synthesized t-test and Fold Change (FC) evaluation were to help identify potential metabolites. Metabolites with the highest variable importance in the projection (VIP) score are the most powerful group discriminators, VIP score > 1 are significant [32]. The significantly differential metabolites were ranked using the VIP score (> 1) based on the OPLS-DA model and P < 0.10. The instructions of metabolites identification and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were according to Wang et al.[30].