|  | Protocol | Mean ΔG |
P
| Mean Z(ΔG) | Mean %pairs |
|---|
Real (mouse) |  | -737.98 ± 55.52 |  |  | 60.96 ± 0.28 |
Modification | Swap G4C4 | -734.10 ± 55.08 | 0.0169 |  | 62.11 ± 0.33 |
Randomization | Sh.4-fold | -725.76 ± 54.71 | 9e-15 | -1.41 ± 0.14 | 60.77 ± 0.23 |
|  | Sh.codon | -728.49 ± 55.01 | 6e-10 | -1.04 ± 0.14 | 60.61 ± 0.23 |
|  | Re-sub.K | -733.28 ± 55.15 | 4e-05 | -0.64 ± 0.15 | 61.06 ± 0.24 |
|  | Re-sub.N3 | -734.14 ± 55.20 | 4e-04 | -0.51 ± 0.14 | 61.09 ± 0.24 |
- Means ± SEM are shown, N = 70. P-values for modifications were determined by paired t-tests (μ = Real < Modification) on ΔG. P-values for randomizations were by one-sample t-tests (expected mean (μ) = 0) on Z(ΔG). %Pairs is the proportion of the coding sequence involved in base-pairing interactions. Artificial sequences generated by the first five protocols encode the same protein as the mouse sequence. A brief description of each protocol follows (see Results for details). 'Sh.4-fold': nucleotides at all 4-fold degenerate sites are shuffled. 'Sh.codon': for each amino acid, the synonymous codons are permuted. 'Re-sub.K': synonymous substitutions are reverted back to the rat-mouse common ancestor (rat-mouse common ancestor) state, followed by reallocation of the same number of synonymous point mutations. 'Re-sub.N3': like the previous protocol, except that the nucleotide replacement is also selected at random from the nucleotide distribution at third sites observed in the mouse sequence. 'Swap G4C4': all guanine bases at 4-fold sites are replaced by cytosine, and vice versa.
