Skip to main content
Fig. 1 | Microbiome

Fig. 1

From: Pathogenic functions of host microbiota

Fig. 1

Association between pathofunction abundance and disease. Abundances of genes encoding three pathofunctions, namely, the formation of (i) trimethylamine (cutCD (a), cntAB (b), grdH (c)), (ii) the secondary bile acids lithocholic/deoxycholic acid (baiA-I (d)), and (iii) hydrogen sulfide (dsrAB (e)), were quantified in metagenomic data encompassing patients (red violin plots) suffering from cardiovascular disease (CVD: I, II), type 2 diabetes (T2 diabetes: III–V), obesity (Obese: VI), colorectal cancer (CRC: VII-IX), liver cirrhosis (Cirr: X), and inflammatory bowel disease (IBD: ulcerative colitis (UC) and Crohn’s disease (CD), XI, XII) and compared with healthy controls (white violin plots). In dataset XII results of CD patients are displayed following those from UC patients (on the right), whereas CD patients of dataset XI are not shown due to the low sample size (n = 4). For the key of colors of additional violin plots, see legend at the bottom of the figure. Pathofunction abundance refers to the percentage of total bacteria of a sample carrying respective genes, i.e., relative to the mean abundance of three single-copy housekeeping genes (HKG). Gene expression results (MTX) are displayed relative to the mean expression of those HKG. Black bars in violin plots represent median values. Significant differences (*p < 0.05) and trends (+p < 0.1 and #p = 0.1) between patients and healthy controls are indicated (Student’s t test on log-transformed (log(x + 1)) data, whereas each disease results of generalized linear mixed-effects models using sample origin (dataset) as a random effect are indicated at the bottom of each plot within dashed arrows

Back to article page