On individual genome-wide association studies and their meta-analysis