Since the distribution of the frequency of the most associated allele observed from GWAS is not consistent with an important contribution of multiple rare variants and since the variance attributable to the causal locus would have to be unrealistically high to be detectable by GWAS conducted to date, we conclude that multiple rare variants are unlikely to explain an important proportion of GWAS results, particularly for associations with very common alleles. GWAS of larger sample size will undoubtedly identify more associations and will point to additional regions in the genome for follow-up studies. A recent GWAS of height of >180,000 individuals [6] has identified 180 loci enriched for genes that are connected in biological pathways and that underlie skeletal growth defects. Much larger sample sizes are needed for GWAS of disease than have been currently conducted to achieve the same power afforded to the height study (e.g., ∼50,000 cases and 50,000 controls for schizophrenia [7]).
