TABLE 4.

Contribution of each size class of variants to the θ-values

			θw			θΠ
	Size class	Frequency	Obs.	Exp.	Difference	Obs.	Exp.	Difference
All	Size 1	40	7.44	2.28	5.16	0.66	0.12	0.54
Size 1–10%	10	1.86	4.80	−2.94	0.46	1.25	−0.79
Size 10–25%	4	0.74	2.04	−1.30	0.94	1.79	−0.85
Size 25–75%	11	2.05	2.51	−0.46	5.19	3.67	1.52
Size 75–99%	1	0.19	0.65	−0.46	0.03	0.46	−0.43
Total	66	12.27	12.27	7.29	7.29
Africans	Size 1	31	7.29	2.82	4.47	1.55	0.36	1.19
Size 1–10%	9	2.12	2.35	−0.23	0.97	0.69	0.28
Size 10–25%	2	0.47	2.81	−2.34	0.52	1.86	−1.34
Size 25–75%	9	2.12	3.19	−1.07	4.18	3.80	0.38
Size 75–99%	0	0	0.82	−0.82	0	0.51	−0.51
Total	51	11.99	11.99	7.22	7.22
Asians	Size 1	7	1.63	1.30	0.33	0.33	0.30	0.03
Size 1–10%	4	0.93	1.40	−0.47	0.54	0.86	−0.32
Size 10–25%	1	0.23	1.10	−0.87	0.32	1.52	−1.20
Size 25–75%	11	2.56	1.42	1.14	5.07	3.23	1.84
Size 75–99%	1	0.23	0.36	−0.13	0.05	0.40	−0.35
Total	24	5.58	5.58	6.31	6.31
Europeans	Size 1	8	1.88	1.38	0.50	0.40	0.35	0.05
Size 1–10%	2	0.47	1.15	−0.68	0.24	0.67	−0.43
Size 10–25%	2	0.47	1.38	−0.91	0.62	1.79	−1.17
Size 25–75%	11	2.59	1.57	1.02	5.31	3.65	1.66
Size 75–99%	2	0.47	0.40	0.07	0.38	0.49	−0.11
	Total	25	5.88	5.88		6.95	6.95

The observed θ_w and θ_Π were obtained by Watterson's θ_w = K/a_n and Tajima's θ_Π = ∑Π_ij/(n(n − 1)/2), respectively. The expected θ_w and θ_Π were obtained from θ_w = K/a_n, where K is the expected variant sites in each class, and θ_Π = θ∑(n − i)/(n(n − 1)/2), respectively.