Figure 2

Crohn’s patients with recurrence have significant dysbiosis at the time of surgery relative to control patients. ( A ) PCoA plot of weighted UniFrac distances between all surgical biopsies and non-IBD colonoscopic biopsies, with Crohn’s surgical samples (red) sized according to Rutgeerts score at post-operative colonoscopy. A Rutgeerts score of 2 or more indicates recurrence; less than 2, remission. ( B ) Average weighted UniFrac distances (±s.e.m.) between Crohn’s surgical samples and surgical controls were significantly greater in recurrence. The asterisk indicates a significant difference: * P<0.05. ( C ) Heat map showing probability of remission or recurrence based on weighted UniFrac distance to a single surgical control biopsy, binned in increments of 0.1. Each row adds up to 1. Based on a single biopsy per Crohn’s patient, differences in UniFrac distance distributions between remission and recurrence were not consistently significant, as demonstrated by ( D ) random selection of a single surgical biopsy from each of the six Crohn’s patients and determination of UniFrac distance to a randomly selected control biopsy (repeated 10,000 times), and ( E ) examination of all 486 combinations of pairwise comparisons between each surgical biopsy from the six Crohn’s patients and the centroid in ordination space of control biopsies. Dark gray shaded bars indicate a P-value of less than 0.05.