Numbers of case-control pairs required to attain a significance level of α = 1 × 10−7 with 95% power in a single-stage study, assuming a multiplicative genetic model with the indicated relative risks per allele (i.e., homozygote relative risk being the square of the indicated relative risks)