We conducted a case control association study for disease Y, with 500 cases and 500 controls. For SNP X, there are two alleles, A and G. For the cases, we detected a total of 400 A alleles and 600 G alleles. For the controls, there are a total of 300 A alleles and 700 G alleles. In genome-wide association studies, due to the multiple testing burden and in order to reduce the chance of false positive detections, we usually use 5x10^-8 as the threshold for genome-wide significant P values to define positive association.

What sample size would be needed in order to reach genome-wide significance?