Fig. 1: Effect of antibiotic exposure duration on non-responder phenotype.

The table in the center denotes the number of non-responder and responder mice in each treatment duration group. A Experimental design for the duration experiment. Circles denote sampled time points. Time points were considered sampled “during” antibiotic treatment between day 0 and day 2, 4, 8, and 16, respectively, as denoted by orange shades. B Relative abundance of phyla on the last day of antibiotics treatment. The control panel is an average over all untreated controls from all time points. Only phyla with a relative abundance of at least 0.1% are shown. Each barchart denotes means from at least two samples and white insets are the sample size used for each barchart. C Percentage of mitochondria and chloroplast sequences in 16S amplicon data relative to antibiotic treatment. Colors: red—controls not treated with antibiotics, green—non-responders, blue—responders. D Principal coordinate analysis (PCoA) of samples during and after antibiotic exposure (n = 143 samples with >10,000 reads per sample, day ≥ 0). Ellipses denote 95% confidence intervals from a Student t-distribution. Each point denotes a sample. ASV abundances were rarefied to 10,000 reads for each sample and percentages in brackets denote the explained variance. Samples with less than 10,000 reads per sample were not included in the analysis. E Dynamics of amplicon sequence variants (ASVs). Gained ASVs are variants that were not present before antibiotics treatment but are present after. Similarly, lost ASVs were present before treatment but not after, and persistent ASVs were present before and after. Stars denote significance under a Mann–Whitney U test: *p < 0.05, **p < 0.01.