. 2009 May 15;5(5):e1000477. doi: 10.1371/journal.pgen.1000477

Table 1. The table shows the power for each chip with a sample size of 2000 cases and 2000 controls and a relative risk at the causal SNP of 1.3 using a p-value threshold of 5×10⁻⁷.

Chip	Chip SNP Tests	MultiMarker Tests	IMPUTE
Affy100 k	0.178	0.212	0.242
Affy500 k	0.363	0.378	0.450
Illu300 k	0.392	0.424	0.467
Illu610 k	0.439	0.455	0.488
Illu650 k	0.443	0.458	0.492
Affy6.0	0.420	0.433	0.478
Illu1M	0.457	0.461	0.493
Complete	0.499	0.499	0.499

Three different methods of analyzing the genotype data from each chip are shown: (a) testing just the SNPs on each chip, (b) using MultiMarker Tests in addition to the tests at each chip SNP, and (c) carrying out imputation using IMPUTE and testing all imputed SNPs in addition to those on each chip. The last line of the table shows the power that woud be obtained using the ‘Complete’ chip.

Table 1. The table shows the power for each chip with a sample size of 2000 cases and 2000 controls and a relative risk at the causal SNP of 1.3 using a p-value threshold of 5×10−7.

Table 1. The table shows the power for each chip with a sample size of 2000 cases and 2000 controls and a relative risk at the causal SNP of 1.3 using a p-value threshold of 5×10⁻⁷.