Two-stage study designs for analyzing disease-associated covariates: linkage thresholds and case-selection strategies

Table 3 Power estimates for study designs

		Proportion of reps. meeting:
		Linkage threshold		Association threshold (avg. no. SNPs analyzed)^c		Avg. no. cases analyzed		Overall power
		Stringent^a	Loose^b	Stringent	Loose	Stringent	Loose	Stringent	Loose
9 (186)	A	0	0.14	0 (n/a)	0 (144)	1500	1500	0	0
	B	0	0.14	0 (n/a)	0 (144)	983	987	0	0
	C	0.13	0.64	0.23 (48)	0.25 (145)	127	136	0.03	0.16
	D	0.83	0.96	0.01 (48)	0 (147)	1500	1500	0.01	0
	E	0.2	0.65	0.30 (48)	0.20 (148)	129	121	0.06	0.13
11 (389)	A	0.06	0.24	1 (52)	1 (147)	1500	1500	0.06	0.24
	B	0.06	0.24	1 (52)	1 (147)	988	985	0.06	0.24
	C	0.11	0.62	1 (48)	0.92 (147)	258	273	0.11	0.57
	D	0.98	1	1 (44)	1 (146)	1500	1500	0.98	1
	E	0.23	0.56	1 (46)	1 (147)	281	290	0.23	0.56
18 (269)	A	0.21	0.64	1 (28)	1 (97)	1500	1500	0.21	0.64
	B	0.21	0.64	1 (28)	1 (97)	998	988	0.21	0.64
	C	0.53	0.91	1 (28)	1 (96)	372	390	0.53	0.91
	D	0.73	0.99	1 (27)	1 (95)	1500	1500	0.73	0.99
	E	0.44	0.88	1 (27)	1 (95)	370	382	0.44	0.88

^aStringent linkage thresholds are LOD score ≥ 1.0, OSA p-value ≤ 0.05, 10-cM linkage region.
^bLoose thresholds are LOD score ≥ 0.5, OSA p-value ≤ 0.5, 40-cM linkage region.
^cThe proportion of replicates meeting the Bonferroni-corrected association threshold, average number of cases and number of SNPs analyzed were calculated across replicates meeting the linkage threshold.

ISSN: 1753-6561