Table 3.

Number of Samples Required by Admixture Mapping versus Linkage and Direct Association Studies to Detect Known Risk Alleles^[Note]

		Increased Risk		Frequency in(%)		Increased Risk for		No. of Samples Requiredfor 80% Power in
Locus (Allele)	Phenotype	Due toHeterozygosityfor Risk Allele(ψ1)	Due toHomozygosityfor Risk Allele(ψ2)	Europeans	WestAfricans	1 European-AncestryAllele	2 European-AncestryAlleles	AdmixtureMapping	HaplotypeMapping	LinkageMapping
CTLA4 (Ala allele in Thr17Ala)a,b	Type I diabetes	1.26	1.74	38	21	1.04	1.08	36,144	2,557	233,169
INS (class I allele in VNTR)c,d	Type I diabetes	2.30	2.86	71	23	1.48	2.19	974	448	8,203
DRD3 (Ser allele in Ser9Gly)b,e	Schizophrenia	1	1.12	67	12	1.01	1.05	346,816	265,999	380,983,674
AGT (Thr allele in Thr235Met)f,g,h	Hypertension	1.12	1.31	42	91	.93	.87	16,034	11,332	4,941,111
PPAR-γ (Pro allele in Pro12Ala)b,i	Type II diabetes	1.3	1.7	85	100	.97	.93	62,134	21,297	18,151,737
CTLA4 (Ala allele in Thr17Ala)c,j,k	Graves disease	1.32	1.80	38	21	1.05	1.10	28,861	2,041	157,555
PRNP (Met allele in Met129Val)c,l,m	CJD susceptibility	1.88	3.57	72	56	1.11	1.23	9,081	422	7,666
APOE (E4 allele)c,n,o	Alzheimer disease	4.2	14.9	14	30	.76	.57	1,165	71	316
F5 (Leiden allele)c,p,q	Venous thrombosis	7.83	80	4	0	1.27	1.62	1,156	134	457
IBD5 (A allele in IGR2096a_1 A/C)r,s,t	Inflammatory bowel disease	1.38	2	35	0	1.13	1.30	4,596	3,918	565,369
KCNJ11 (Lys allele in Glu23Lys)u	Type II diabetes	1.12	1.47	34	3	1.02	1.05	43,312	22,466	15,589,550
HLA DR2 (DRB1*1501)v	Multiple sclerosis	2.7	6.7	11	0	1.19	1.40	2,498	678	16,047
ABCB1 (C allele in C3435T)w	Epilepsy treatment	1.47	2.66	50	10	1.20	1.50	1,985	969	30,623
GNB3(T allele in C825T)x	Obesity (BMI >27)	1.98	3.59	30	81	.75	.55	1,055	602	15,704
β-globin (Val allele in Glu6Val)y	Sickle-cell disease	1	1,000	0	6	.22	.22	92	5	14

Note.— To estimate the increased risk due to 1 or 2 European ancestry alleles, we used the frequencies of the risk alleles in European and West Africans and the increased risk due to one or two copies estimated in European Americans. For calculating the power of linkage analysis and admixture mapping, we assumed fully informative maps, and assumed 300,000 independent hypotheses for direct association studies. The first nine lines in the table show associations with complex disease identified by Hirschhorn et al. (2002), Lohmueller et al. (2003), and K. Lohmueller (unpublished data) that were significant in meta-analysis or reproducible in 75% of follow-up studies (with the caveat that their frequencies were available in Europeans and Africans). The odds ratios are calculated from follow-on studies, where increased risk due to heterozygosity was estimated using the odds ratio for the risk allele rather than the heterozygous genotype. Lines 10–14 show less well-established associations with complex disease, and line 15 shows a Mendelian disease.

^aOsei-Hyiaman et al. 2001.

^bLohmueller et al. 2003.

^cHirschhorn et al. 2002.

^dPermutt and Elbein 1990.

^eCrocq et al. 1996.

^fRotimi et al. 1996.

^gNakajima et al. 2002.

^hK.E.L., unpublished data.

ⁱAltshuler et al. 2000.

^jUeda et al. 2003.

^kDonner et al. 1997.

^lMead et al. 2001.

^mSoldevila et al. 2003.

ⁿFarrer et al. 1997.

^oCorbo and Scacchi 1999.

^pRosendaal et al. 1995.

^qRees et al. 1995.

^rRioux et al. 2001.

^sGiallourakis et al. 2003.

^tM.J.D., unpublished data.

^uD.A., unpublished data.

^vBarcellos et al. 2003.

^wSiddiqui et al. 2003.

^xSiffert et al. 1999.

^yHill et al. 1991.