In summary, we explored the genetic architecture of AD in the AA population via the analysis of common variants. Our results support the notion that AD is a highly polygenic trait, i.e., there exist many risk variants conferring small or moderate effects. The limited sample size is a major issue with respect to the goal of identifying the causal variants. However, sample recruitment is expensive and time-consuming. Some alternative ways to boost statistical power of GWAS data analysis can be considered. First, accumulating evidence suggests that different complex human traits are genetically correlated, i.e., multiple traits share common genetic bases, which is known as pleiotropy. It is a promising direction to exploit the pleiotropy between AD and other psychiatric disorders by combining multiple GWAS sets, because the sample size can be effectively increased. Our findings also show that, as would reasonably have been predicted, SNPs do not equally contribute to the AD risk (nor do chromosome or functional units; chromosome 4 (physical region) and regions within 20kb of genes (functional region) contribute more). Thus, incorporating biologically relevant information could be
