Correcting for multiple testing in genetic association studies: the legend lives on