Fig. 3: Significant metabolic changes and associated pathways upon glucose treatment in deep samples.

Panel (A) summarizes the results of Orthogonal Projections to Latent Structures—Discriminatory Analysis (OPLS-DA), performed independently for control and glucose treatments discriminating sampling depth across each metabolite datasets. Results are represented using a DiVENN diagram¹⁰¹ to highlight unique and shared discriminatory metabolites identified for control and treatment groups. Each dot represents a metabolite which is colored based on the abundance; surface abundant (surface > deep: green) and deep abundant (surface <deep: brown). Shared discriminatory metabolites with differences in their abundances across two analyses (surface > deep for control but surface <deep for glucose, vice versa), were colored yellow. The abundance of unique discriminatory metabolites (for the control and glucose-amended group) detected with NMR and GC MS methods are shown in the heatmaps. Each metabolite in the heatmap was scaled using a z-score to enhance the visualization of abundance distribution. Using ¹H NMR, the consumption of ¹³C labeled glucose was tracked separately, and the ¹³C enriched metabolites identified were cross referenced to the above depth-discriminatory metabolites, irrespective of the treatment. These metabolites are denoted with a star. Panel (B) shows the differential pathway shifts for a selected set of pathways based on metabolites that showed an absolute log2 fold change in abundance greater than 1 between treatments, for aggregated timepoints. Bars attest to the number of matched metabolites to the respective KEGG reference pathway⁶³, colored based on their comparative abundance.