Effect sizes reported for SNPs associated with complex traits have generally been small. Thus, the power of a single GWAS to find SNPs showing unequivocal evidence of association is low. A likely reason for the small effect sizes observed for each individual SNP is genetic heterogeneity - that is, different individuals possess susceptibility alleles at different SNPs, or even different genes. This will greatly reduce the power to detect association with any single SNP. However, if the genes in question influence etiology through a common biological pathway, tests based on the pathway as a whole should give increased power to detect association, even in the absence of any genome-wide significant single SNPs, as well as being more replicable between studies.
