Fig. 5

Functional enrichment in core genes versus sequence diversity in coding or flanking intergenic sequences. a Workflow for identifying genes with high or low sequence diversity. For a given gene and species, frequencies of individual coding, 3′ intergenic (3’IG), and 5′ intergenic (5’IG) variants were computed, and entropies of the three variant type-specific frequency distributions were computed as measures of sequence diversity. For a given entropy measure, genes in the top and bottom 5% were classified as “diverse” or “conserved”; in the case of coding sequence entropy, 5 and 95% percentiles as a function of gene length were used instead, estimated through quantile regression. b Functional enrichment in genes classified as most diverse or conserved by either coding, 5’IG, or, 3’IG sequence entropy. Only COGs with positive mean log2 odds ratio (LOR) across the 12 species for at least one entropy measure are shown