Table 1.

Results of the simulations: Percentage of quartets in which significant windows were found and C(k)₁₂ values (average ± s.e.m.) for those windows. The values of μ refer to the rate applied to the region shown in Figure 1 (amino acids 76 – 125 in the simulated proteins), while the other regions of the sequences (amino acids 1 – 75 and 126 – 200) were kept evolving at a basal rate μ = 0.5. α is the shape parameter of the gamma distribution. The percentage of significant simulations was compared between the controls and the quartets with increased rates using the chi-square test. The average C(k)₁₂ values were compared using the t test. In bold, significant results. p values are detailed

	α = 0.2	α = 0.7	α = 1.5	α = 3
μ = 0.5 (control, no increase)	2% 0.881 ± 0.146	3% 0.664 ± 0.208	2% 0.715 ± 0.238	3% 0.831 ± 0.226
μ = 1 (2× increase)	2% 0.613 ± 0.306	3% 0.583 ± 0.218	3% 0.914 ± 0.124	3% 0.909 ± 0.102
μ = 2.5 (5× increase)	5% 0.814 ± 0.133	13% (p = 0.012) 0.998 ± 0.070 (p = 0.037)	12% (p = 0.008) 0.808 ± 0.075	14% (p = 0.007) 1.010 ± 0.096
μ = 5 (10× increase)	10% (p = 0.020) 1.048 ± 0.220	35% (p < 0.001) 1.188 ± 0.045 (p = 0.002)	55% (p < 0.001) 1.328 ± 0.114 (p = 0.001)	51% (p < 0.001) 1.374 ± 0.049 (p = 0.006)