Fig. 5

The posterior estimated predictive distribution of relative abundances at family level in ACB and BCA sequence. Top of the 20 most abundant families. Points are the posterior population mean. The bars are the 89% prediction intervals from the posterior distribution. Inter-subject variation is not included. Baseline [high-protein] (diet C), High-insoluble Fiber (diet B), Hydrolyzed (diet A) and Washout [high-protein] (diet C)