Fig. 6
From: Characterization of the gut microbiome in a porcine model of thoracic spinal cord injury
Functional inference analysis comparing SCI and control specimens using PICRUSt2. A) PCA Biplot. A biplot is a combination of a samples PCA plot and a loadings plot that shows how strongly each characteristic influences a principal component. Boxed names are the loadings. Ellipses represent the default 95-level assuming a multivariate t-distribution. B-E) Longitudinal characterization of bacterial abundance of the 4 pathways with the strongest negative (B, C) and positive effect (D, E) size. The individual median CLR values are plotted along with the associated loess curves (Local Polynomial Regression Fitting) with confidence intervals in gray. Red = SCI, Blue = Control