Table 3.

Sample Sizes Necessary to Detect an Association between a Marker Locus and a Trait-Influencing Locus Assuming Different Values for the Trait-Locus Allele Frequencies, Locus Effect Sizes, and LD Strength^[Note]

			Necessary Sample Size for Assumed Model ofInheritance and Assumed Type I–Error Rate
			Dominant		Recessive		Additive
Trait-LocusAlleleFrequency	LD (D′)	Locus-SpecificHeritability(LocusEffect)	.05	.00001	.05	.00001	.05	.00001
.10	.75	.10	150	425	1,001	2,856	140	394
.10	.50	.10	330	941	2,215	6,318	305	871
.10	.25	.10	1,297	3,700	8,701	24,819	1,197	3,415
.10	.75	.20	77	220	974	2,778	71	203
.10	.50	.20	172	482	2,154	6,143	156	444
.10	.25	.20	650	1,855	8,458	24,126	602	1,716
.25	.75	.10	56	160	105	302	48	135
.25	.50	.10	125	357	229	655	106	305
.25	.25	.10	500	1,427	885	2,524	417	1,191
.25	.75	.20	29	86	50	146	24	70
.25	.50	.20	67	193	110	315	113	157
.25	.25	.20	270	767	417	1,190	216	615

Note.— The associated marker-locus allele was assumed to have a frequency of .25, sampling was assumed to involve the upper and lower 10th percentiles of the trait distribution, and the power was assumed to be 80%. The entries reflect the number of case and control subjects (which are assumed to be sampled equally in number).