Fig. 3: Metabolic signatures of enhancing versus non-enhancing tumours versus brain.
From: Blood-brain barrier disruption defines the extracellular metabolome of live human high-grade gliomas
a (i) 162 metabolites present in at least 40/44 catheters were ranked according to the enhancing tumour-versus-brain fold change in each patient. The rank order of each metabolite in each 2-catheter tumour/brain comparison (e.g., catheter X versus Z) is conveyed as a heat map from 1 (orange, enhancing tumour) to 162 (green, brain). Metabolites are listed based on the average of ranks of enhancing vs. brain across nine cases. (ii) Normalised enrichment scores (NES) are shown based upon enrichment analysis for the top 35 enhancing tumour (E) and brain (B)-associated metabolites in the ranked metabolite of enhancing tumour vs. brain (Cath. X/Z) (bold: significant at FDR < 0.05; Complete FDR details available in Supplementary Table 2). See Supplementary Fig. 1 for graphical depiction of how enrichment analyses are performed by repurposing the Gene Set Enrichment Analysis software. b Using the same method as in 3 A(i), the rank order of each metabolite in enhancing-versus-non-enhancing catheter metabolite lists (Catheter X and Y) is depicted as a heat map in the 7 patients for whom both enhancing and non-enhancing catheters were present. Metabolites are listed based on the average of ranks of enhancing versus non-enhancing tumour across seven patients. (ii) As in 3 A(ii), using enhancing versus non-enhancing tumour metabolite ranked lists and metabolite sets. NES are shown for the enrichment of each E vs. NE ranked list for other E vs. NE metabolite sets (bold: FDR < 0.05; Complete FDR details available in Supplementary Table 3). c (i) As in 3A-B, the ranked order of metabolites in non-enhancing versus brain metabolite lists is shown as a heat map. Metabolites are ranked based on the average rank of NE tumour to brain across 7 patients. (ii) As in prior enrichment analyses, with non-enhancing versus brain ranked lists and metabolite sets and their respective NESs across patients (bold: FDR < 0.05; Complete FDR details available in Supplementary Table 4).
