Fig. 3
From: Diet and gut microbiome interactions of relevance for symptoms in irritable bowel syndrome
Co-inertia analysis associates microbiota profiles with dietary profiles, together with gas metabolism and symptom severity. a Co-inertia scatter plot with the training set (n = 79) including both metagenomic and dietary data, and the test set (n = 122) for which only dietary or metagenomics data were available. Individual coordinates for the test set were computed from their PCoA coordinates with the co-inertia model. Colors indicate the data source (diet or metagenomic). b Heatmap of Spearman’s correlations between the first seven co-inertia axes and clinical, gas metabolism, gene richness and dietary data. The color indicates the strength of the correlation. Black squares indicate missing data. v, validation set; t, training set; m, microbiota; f, food. c Scatter plots of the relation between predicted (linear regression) and observed data for H2/CH4 ratio (r = 0.53, p < 0.05), meat/plant ratio (r = 0.69, p < 0.05) and symptom severity (r = 0.28, p < 0.05) (based on the first 5 co-inertia components). The color indicates the validation set source, whereas the shape of the points indicates the training set source (dietary or metagenomic). IBS symptom severity groups (healthy, mild, moderate, and severe) were coded from 1 to 4, respectively