Fig. 2

Short-term calorie restriction alters the gut microbiota in mice. a PCoA based on the Bray-Curtis distance. b Clustering of the gut microbiota between different groups calculated with multivariate analysis of variances (MANOVA) test using the first 25 principal components (PCs) (accounting for 80.73% of total variations) of PCoA based on Bray-Curtis distance. ***P < 0.001. c The rank of the 50 OTU-level phylotypes identified as key variables for differentiation between the gut microbiota of the CR group and that of the NC group by random forests. Day 84 samples from the NC and CR groups were included for discrimination. OTUs were ranked in descending order of important scores, which were determined by estimating the mean decrease in the accuracy (MDIA) when a specific OTU was removed from the set of predictors. Data are shown as the mean ± s.e.m. d Heat map of the 50 key OTUs. The colour of the spots in the panel represents the relative abundance (normalized and log2-transformed) of the OTU in each sample. The OTUs were organized by Spearman’s correlation analysis based on their relative abundances. The genus-level taxonomic classifications of the OTUs are shown on the right. n = 10 for both groups at all time points except n = 9 for the CR group at day 7 and day 28 and n = 8 for the NC group at day 28