Overview of analysis pipeline and example results.

<p>(A) shows the SNV calling and analysis workflow, consisting of an optional pre-processing step, which splits the computation load into subsets of similar size based on the genome coverage, the main SNV calling step and further post-processing of the raw output, which can be tailored according to the aim of the analysis. (B) shows the Principal Coordinate Analysis projection of a pairwise distance between oral samples, based on population SNVs, which clearly separates strain populations in tongue dorsum samples from those in supra-gingival plaque samples. (C) shows the tracking of the individual SNV frequencies within an individual over a period of 384 days. Each line represents one variant position and the respective colour encodes the amount by which the allele frequency of that position changed over time; red represents stable variants that maintain their frequency while in blue are positions which dramatically change their frequency in the population. Only a small number of positions vary over the measured period, with most remaining at approximately the same population frequency, suggesting great stability of strain populations within the individual.