Synonymous substitution rate.

Ka and Ks represent the frequencies of nonsynonymous (N) and synonymous (S) substitutions, respectively, with the unit of per nucleotide, and were calculated according to the following equations: ,

The probabilities of synonymous and nonsynonymous substitutions were 23.1% and 67.9%, respectively, based on the length of the ORF region of 4,148,273 bp (GenoBase, W3110, repeated sequences and overlapping regions among the ORFs were subtracted). Thus, the values applied here were 0.231 and 0.769 for the probabilities of synonymous and nonsynonymous sites, respectively, and 4,148,273 for the length of total ORFs in the E. coli genome. The synonymous substitution rate was Ks divided by the corresponding generation. The number of nonsynonymous substitutions from Anc to 43B, i.e., 10, was significantly larger than the expected number based on the commonly observed synonymous substitution rate of 10−10 (see Discussion) and the above probabilities (one tailed tests, P<0.01).