Fig. 2: Transmission dynamics within the Maela refugee camp.

a, Top: the distribution of pairwise geographic distances between 411 different households versus the number of intermediate transmission events as inferred using the modified TransCluster algorithm. The median and interquartile range are given by the horizontal lines, with the whiskers indicating the largest and smallest values excluding those outside 1.5 times the interquartile range. Bottom: the distribution of estimated intermediate transmission events within households. b, A map of the Maela refugee camp, with inferred direct transmission links overlaid. Roads are shown in white. The direction of transmission is not estimated. Blue lines indicate transmission links that would typically be inferred using a representative genome per sample, while red lines indicate additional links that were found using PDS. c, A representative mother-child pair indicating how transmission direction was inferred. Coloured circles indicate the serotypes present, with PDS data available for those coloured in darker shade. Black lines indicate close transmission links inferred using the TransCluster algorithm, with the vertical red line indicating the time the child was one-year old. d, The distribution of the direction of transmission between mother and child split by whether the transmission event occurred before or after the child turned one. e, A schematic demonstrating that we would expect to see an elevated rate of polymorphic sites (represented by blue and red variants) among close transmission pairs. f, The distribution of the number of shared polymorphic sites in 3,663 potential transmission pairs involving an estimated 0, 1 or ≥2 intermediate hosts. The elevated number of variants involving 0 intermediates hosts indicates a mean bottleneck size ≥1. The median and interquartile range are given by the horizontal lines, with the whiskers indicating the largest and smallest values excluding those outside 1.5 times the interquartile range.