at a considerable distance from the associated common variants. Second, it is now known that rare variants contribute to common diseases, and that cases that carry the rare high-penetrant contributors to disease often have “typical” clinical presentations [19]–[21]. On balance, therefore, our results suggest that even though the apparent impact of common variants is only modest for many traits [6],[22],[23], this impact may have been systematically overestimated [24]. It is worth emphasizing that the alternative explanation provided here makes clear, testable predictions. As noted, in a model of synthetic associations, regions that show significant effects for common variants will often consistently show significant residual independent effects after the effect of the most important variant has been accounted for. Second, since rare variants are much more likely to be population specific, synthetic associations are expected to be inconsistent across population groups. In fact, a number of recent studies have confirmed differences in effect between populations [24]–[35]. Table 1 lists variants from these studies in which the point estimate for a follow-up study in a separate population fell outside the 95% confidence interval for the odds ratio of the original study. This includes 13 variants and odds ratios with confidence intervals for
