Extended Data Fig. 2: Verification of resistance calling pipeline as well as the distribution of resistance calls in mothers and infants.

(a) The number of resistance calls identified in 584 samples which consisted of only a single pneumococcal lineage and were sequenced using PDS and via single colony picks in Chewapreecha et al., 2014. The high correspondence between the two methods suggests PDS has a low false positive rate. (b) The number of samples found to be either resistant or susceptible to each antibiotic class for both mothers and infants. Resistance was determined by running the CDC pneumococcal resistance pipeline on the deconvoluted lineages output by the mGEMS pipeline. The individual lineage calls were collapsed to the sample level so that a sample was called as ‘resistant’ if resistance was observed in any of its lineage.