. 2008 Feb 1;82(2):453–463. doi: 10.1016/j.ajhg.2007.11.003

Table 2.

Size and Power of Tests at Level 0.05

			Eigenstrat with F_ST			pMatch with F_ST			fMatch with F_ST
Statistic	Design	No. of Markers	0.01	0.03	0.1	0.01	0.03	0.1	0.01	0.03	0.1
Size

	Two Populations
		96	.069	.106	.211	.062	.100	.202	.065	.101	.206
		386	.055	.061	.085	.044	.045	.051	.047	.049	.054
		1536	.052	.054	.055	.046	.045	.045	.047	.047	.046
		6144	.053	.052	.051	.044	.045	.045	.047	.047	.046
		12000	.053	.052	.052	.044	.044	.045	.047	.045	.047
		24000	.053	.050	.051	.043	.042	.041	.046	.046	.045
	Gradient
		96	.069	.109	.221	.049	.067	.109	.061	.097	.201
		386	.054	.058	.071	.043	.046	.048	.048	.052	.063
		1536	.052	.051	.050	.045	.044	.044	.047	.047	.046
		6144	.052	.051	.050	.045	.045	.044	.046	.045	.047
		12000	.052	.052	.052	.044	.044	.045	.047	.047	.047
		24000	.052	.052	.052	.044	.045	.044	.047	.046	.046

Power

	Two Populations
		96	.783	.710	.683	.693	.635	.620	.754	.685	.659
		386	.767	.701	.682	.682	.632	.621	.735	.673	.653
		1536	.766	.702	.677	.683	.635	.622	.736	.674	.653
		6144	.765	.694	.676	.684	.630	.623	.735	.671	.653
		12000	.765	.697	.676	.684	.633	.624	.735	.673	.653
		24000	.763	.696	.677	.684	.632	.624	.734	.671	.653
	Gradient
		96	.939	.917	.833	.886	.872	.804	.922	.900	.814
		386	.924	.891	.796	.877	.857	.782	.902	.869	.775
		1536	.917	.876	.774	.876	.850	.775	.894	.856	.754
		6144	.913	.874	.768	.873	.849	.773	.892	.853	.747
		12000	.915	.873	.771	.874	.849	.774	.891	.850	.749
		24000	.912	.874	.768	.873	.849	.771	.892	.852	.747

Columns depict the results as F_ST varies (0.01, 0.03, and 0.1) in the candidate markers. Results are shown for two scenarios: a two-population mixture and a nine-population gradient. For the size, the expected number of p values smaller than 0.05 is 50.