Fig. 5

Transplantation of p-Cresol-induced microbiota impacts microbial composition, increases p-Cresol production and selective bacterial taxa predict social deficits and stereotypies. A–C α-Diversity as measured by observed number of ASV (A), Shannon’s (B), and Pielou evenness’ (C) indexes (n = 14 FMT^Control, n = 12 FMT^p-Cresol); Kruskall-Wallis pairwise test: p > 0.05. D β-diversity as measured by Aitchison distances from control and p-Cresol groups based on 16S rRNA gene sequencing (n = 14 FMT^Control, n = 12 FMT^p-Cresol). Group differences were tested by pairwise PERMANOVA, p = 0.001. E Synthetic cladogram presenting dysregulated ASV and bacterial taxa increased (green) or depleted (blue) upon FMT^p-Cresol identified by ANCOM. Symbol size reflects the value of the corresponding CLR. F, G Pseudocounts plots from selected ASV increased (F) or depleted (G) upon FMT^p-Cresol identified by ANCOM; Mann-Whitney U-Test: **p < 0.01, ***p < 0.001. H, I Microbial composition prediction of behavioral scores after FMT (n = 14 FMT^Control, n = 12 FMT^p-Cresol). ASV best predicting social interaction deficits (H) and stereotypies (I), as identified by random forest analysis. Only ASV contributing > 1% accuracy in behavioral scores prediction are presented. ASV identical to or ASV pointing towards the same species as ASV identified by ANCOM as increased or depleted in FMT^p-Cresol as compared to FMT^Control are labeled in green and blue respectively. J Fecal levels of p-Cresol 3 weeks post-FMT (n = 15 FMT^Control, n = 15 FMT^p-Cresol); Mann-Whitney U-test: ****p < 0.0001. A–D, F, G Data are presented as violin plots featuring frequency distribution of data, median (dashed line) and quartiles (dotted line). J Data are presented as dot-plots featuring means ± SD