Table 2 Model comparisons of structure categorizations based on the Deviance Information Criterion (DIC), where the smallest value is considered the best model fit. For simplicity, only models which are an improvement over the model over the best secondary structure model (Table 1) are shown, with the exception of the Structured and IDR model for E. coli. H: helix. E: sheet. C: coil. Superscripts S and D indicate if the secondary structure predictions include predictions from structured regions S or IDRs D, respectively, i.e. S = H ^SE^SC^S. R = H ^DE^DC^D

From: Quantifying shifts in natural selection on codon usage between protein regions: a population genetics approach