Fig. 3: Geographic differences in microbiota development.
A Longitudinal analysis of alpha diversity. Shannon index was calculated at genus level. Cross-sectional comparisons were performed using ANOVA with post-hoc Tukey tests. Longitudinal comparisons were performed using mixed-effects regressions with week of life as a covariate and study ID as a random effect. Pairwise longitudinal comparisons between countries were FDR corrected. See Fig. 1 legend for box plot parameters. B Proportion of variation in microbiota composition associated with country. R2 and statistical significance were determined by PERMANOVA using genus-level unweighted Bray–Curtis distances. C Longitudinal plot of mean genus abundances. Genera are displayed if they were present with a mean relative abundance of ≥5% in at least one country at one or more timepoints. D Cross-validation accuracy of Random Forests for prediction of country. Genus relative abundances served as input for each model. Median out-of-bag accuracy (proportion correctly assigned) and interquartile range across 20 iterations of 5-fold cross-validation are displayed. A random subset of 50 samples per country was used for each iteration. E The 10 most important genera selected by Random Forests for discriminating infants by country at the time of the first dose of ORV. Mean cross-validation importance scores based on Gini index are depicted alongside the prevalence and mean abundance of the corresponding genera in each country. C, country; CV, cross-validation; IND, India; MLW, Malawi; ns, not significant; †, +2 weeks in UK due to later vaccination schedule; *p < 0.05; **p = 0.001; ***p < 0.0005.
