Fig. 4

Correlation between effective population size, genome size and pan-genome size. a. Genome size vs. Ne. Genome size for a given species was calculated as the average across all sequenced strains, and Ne was computed from the average dS/dN ratio determined for common set of universally distributed genes (as described in Fig. 1 and Methods, Spearman’s rho = 0.30, P < 0.001, PIC correction). Colors indicate the life style of each species: green: free-living, blue: commensal; red: obligate animal pathogen, light green: plant pathogen, purple: obligate intracellular. b Genome size vs. absolute estimates of Ne (Spearman’s Rho = 0.77, P < 0.05). Genome size for a given species was calculated as the average across all sequenced strains, and Ne was computed for the species with known mutation rates using Watterson’s estimator. c Pan-genome size vs. Ne. Pan-genome size for a given species was calculated as the total number of protein families detected normalized by strain number, and Ne was computed from the average dS/dN ratio determined for common set of universally distributed genes (Spearman’s rho = 0.48, P < 10^− 8, PIC correction). d Pan-genome size vs. absolute estimates of Ne