False positive rates for association studies when a definite number of neutral variants are identified only in cases or when variant discovery is carried out in cases and controls.

The false positive rates are displayed for when M = 5 or 10 neutral variants with equal population frequencies of 0.001, 0.002, and 0.005 were discovered in cases (upper panel) or in both cases and controls (lower panel) for N = 100, 200, 500, and 1,000 cases with equal number of controls. It is assumed that each rare variant resides on a separate haplotype. Analyses were carried out using the Cochran–Armitage test for trend (see Methods). The false positive rates were evaluated for an α = 0.05 and based upon 100,000 replicates.