Table 1.

Hypothetical WGDs, posterior mean of duplicate retention rate (q), and the Bayes Factor (K) to compare the likelihood of q = 0 (H₀) to the likelihood of q > 0 (H₁) using the Savage-Dickey density ratio.

Hypotheses	Relaxed branch-specific model		Critical branch-specific model
	$\overline{q}$	K	$\overline{q}$	K
WGD1	0.001	1272.297	0.029	2.693
WGD2	0.028	2.649	0.058	0.393
WGD3	0.352	0.047**	0.263	0.061**
WGD4	0.170	0.197*	0.193	0.094**
WGD5	0.156	0.149*	0.006	128.480
WGD6	0.020	21.026	0.000	6919.539
WGD7	0.127	0.166*	0.142	0.135*
WGD8	0.025	12.864	0.027	6.007

^**K < 1/10 or K < 0.1, strong evidence against H₀;

^*K < 1/10^0.5 or K < 0. 3162 substantial evidence against H₀;

K < 1, H₁ supported, not worth more than a bare mention; K > 1, H₀ supported.