The closely related serotypes with less nucleotide divergences had similar hybridization profiles across the tiled regions, so it is possible to assign species (HRVA and HRVB) based only on the hybridization patterns.
As RNA secondary structure is conserved among species and the fitness of structurally disruptive mutations is low, it is reasonable to hypothesize that selection on RNA secondary structure should lead to less nucleotide sequence divergence in the genome.
We identified 123 unique, full-length toxin-coding sequences, which cluster into 78 groups with less than 1% nucleotide divergence, and 2,879 unique, full-length nontoxin coding sequences.
Toxin sequences were clustered into groups with less <1% nucleotide divergence in their coding sequences, and duplicate nontoxin sequences were eliminated following alignment of the final transcripts with NGen.
Members of the genus are often morphologically and ecologically indistinguishable and species can differ by less than 0.1% nucleotide divergence at nuclear loci [ 35] but extensive divergence is observed in mate-recognition related behavioral characters [ 36, 37].
To better group the homologs, genes with a nucleotide divergence less than 0.3 were defined as a family.
To avoid biased evolutionary analyses due to poor resolution, we removed all but one representative for all groups in which Wolbachia strains showed less than 0.2% pairwise nucleotide divergence across the 12 HR genes.
After removal of strains showing less than 0.2% pairwise nucleotide divergence across the 12 HR genes (see Materials and Methods), our final data set consisted of 18 supergroups A and B Wolbachia strains and wBm (supergroup D) as an outgroup.
More distantly related strains (found in other clades of B. cereus s.l. and other Bacillus species) would be surveyed less efficiently because of their greater nucleotide divergence.
