To set a baseline for multiplicity adjustments, we have estimated a genomewide significance threshold of about 7.2 × 10−8 for two–sided tests of single SNPs in the UK Caucasian population. This is similar to the threshold suggested by the International HapMap Consortium [2005], which is reassuring as that study took a complementary approach. They extrapolated near-saturated local regions to the whole genome, whereas we took sparse data over the whole genome and extrapolated to complete saturation. Nevertheless, our estimate is not precise: in contrast to previous approaches we are able to report a confidence interval, which reflects the variation due to the finite number of permutations and random subsampling of SNPs. The half–saturation parameter is about twice the current number of markers, suggesting that the current marker density reflects less than half the total multiplicity in the genome. This does not imply that the marker panels have insufficient coverage, for it has been estimated that up to 80% of the variation in the genome is correlated to current panels [Barrett and Cardon, 2006]. Rather, the statistical cost of obtaining all
