TABLE 2

The probability of hypothetical and published DNA polymorphism surveys under the neutral model

Locus	Sample size	Variable sites	θ_W	θ_π	θ_H	d	P(BM)	No. of occurrences of each derived variant	P(D)	P(H)
Example	10	1	0.35	0.20	1.80	0.010	0.004	[9]	0.430	0.028
Example	10	3	1.06	0.60	5.40	0.010	0.004	[9 9 9]	0.128	0.002
Example	10	3	1.06	0.60	3.62	0.010	0.004	[1 9 9]	0.128	0.018
Example	10	3	1.06	0.60	1.84	0.010	0.004	[1 1 9]	0.128	0.090
Example	10	3	1.06	0.60	0.07	0.010	0.004	[1 1 1]	0.128	0.665
achaete	147	3	0.54	0.87	3.85	0.062	0.084	[84, 120, 141]	0.863	0.027
asense	6	6	2.63	2.20	4.60	0.038	0.014	[1, 1, 1, 4, 5, 5]	0.215	0.077
Pgd	13	14	4.51	5.44	6.06	0.069	0.026	[1(2), 3(3), 4(3), 5(3), 10(2), 11]	0.825	0.298
vermilion	45	6	1.37	0.81	2.55	0.022	0.008	[1(3), 3, 24, 44]	0.160	0.070
Acp26Aa	101	54	10.41	8.07	17.97	0.117	0.044	See Figure 4	0.083	0.043

d is divergence to the outgroup. P(BM) is the probability of a backmutation (materials and methods). P(D) and P(H) are the probability of observing a more negative D and H value under neutrality (α = 0.05, one sided), respectively. Numbers in parentheses indicate the number of sites where the derived variant is observed; i.e., Pgd has two singletons. For Acp26Aa, 4Nc/bp = 0.0725 was used to generate the expected distribution of D and H values. At two of the three polymorphic sites at the achaete locus, the derived variant was inferred to be a loss of a restriction site.