Fig. 7: Interplay between luminal metabolites and blood metabolic responses.
a, Dynamics of luminal metabolites and blood responses. Each block represented the mean concentrations of participants (n = 10), Each variable was normalized using z-score transformation. SR, stachyose/raffinose. b,d, PLSR model correlation loadings for metabolites and blood responses at postprandial 30 and 120 min. Metabolites and associated blood responses were analysed via PLSR models (x (metabolites)/y (blood responses), n = 10). Positively correlated variables are proximal within the correlation circle (angle <0), while negative correlations appear diametrically opposed (angle <180). Variables lacking correlation are situated at angles <90 relative to each other. The x and y axes denote the explained variance percentages by the PLSR factor. c,e, Correlation networks showing significant correlations between x (metabolites) and y (blood responses) at postprandial 30 and 120 min (n = 10), using an FDR-adjusted P value threshold of <0.05 using the Benjamini–Hochberg method. Gastric and duodenal metabolites are indicated with a suffix ‘.g’ or ‘.d’. Standard three-letter abbreviations are used for amino acids.
