Masshunter workstation software for qqq

All raw metabolite data was analyzed using Quantitative Analysis MassHunter Workstation software for QQQ (Agilent Technologies, Santa Clara, CA, USA). Outliers were excluded if the data point was outside the 1.5 × interquartile range (Miller, 1993 (link)). Minitab (Minitab Inc., State College, PA, USA) was used for statistical analysis. Statistical significance between experimental groups (treatments and varieties) was performed using Student’s t test and one-way analysis of variance (ANOVA). False discovery rate (FDR) (Hochberg and Benjamini, 1990 (link)) was used to reduce type I errors in multiple comparisons. Bar plots and line plots were created using Graph Pad Prism 6.0 (GraphPad Software, La Jolla, CA, USA). Hierarchical cluster analysis was carried out in MetaboAnalyst 3.0¹ (Xia et al., 2015 (link)).

Raw metabolite and lipid data were processed and analyzed using Quantitative Analysis MassHunter Workstation software for QQQ (Agilent Technologies, Santa Clara, CA, USA). The level of identification was carried out based on the Metabolomics Standards Initiative (MSI) requirements [43 (link)]. For all measured metabolites (sugars, organic acids, amino acids, and amines), absolute concentrations were determined for all the metabolites except for fructose-6-phosphate and glucose-6-phophate as the linear standard curves for these two metabolites could not be generated for quantification due to very high abundance of these two metabolites in the spike samples. The concentration unit for all the metabolites was expressed as picomole/mg of fresh weight except responses normalized to mg of fresh weight for fructose-6-phosphate and glucose-6-phophate (Table S3). All the metabolites measured at MSI Level 1, as the identification was based on multiple reaction monitoring (MRMs) established using authentic standards (Table S1). For lipids, responses were normalized to mg of fresh weight, FW (Table S4). Although single authentic lipid species were used for each of the phospholipid class, the identification of individual lipid species was based on both the MRM experiment and retention time and was therefore measured at MSI Level 2 (Table S2).