2.3. Data acquisition

Experiments were performed as previously detailed.21 Briefly, 15 μL of serum samples was used for protein precipitation using methanol containing a mixture of internal standards (Table S2). Samples were then mixed, centrifuged, dried under nitrogen and resuspended in 45 μL of 10 mmol/L of ammonium carbonate (pH 10.5)/ACN, 40/60 (v/v). Serum samples were then introduced into a Transcend 1250 LC system (Thermo Fisher Scientific, Les Ulis, France) fitted with a Sequant ZICpHILIC 5 μm, 2.1 × 150 mm (Merck, Darmstadt, Germany) and coupled to a Q‐Exactive mass spectrometer (Thermo Scientific, San Jose, CA) operating in both positive and negative ionization modes, alternatively. In positive ion mode, positive ions are detected while in negative ion mode, negative ions are detected. Quality control (QC) samples composed of an equal amount of each sample were discarded all along the analytical sequence. Data analysis was performed with TraceFinder 3.1 (Thermo Fisher Scientific, Les Ulis, France). Compound identification was carried out by comparing their exact m/z ratio of their corresponding protonated ion [M+H]⁺ in positive ion mode and of their deprotonated ion [M‐H]^‐ in negative ion mode, their retention time and their isotopic pattern to an in‐house chemical library. The obtained data set was cleaned based on several parameters: the coefficient of correlation between serial dilutions of QC samples, the coefficients of variation of the areas of chromatographic peaks of features in QC samples and the ratio of chromatographic area of biological to blank samples QC samples.22 Data and metabolites identifiers are provided in Tables S3 and S4, respectively.