The number of variants that contribute to the disease-risk (D-variants) determines the marginal PAR of each variant in the group, such that a low number of D-variants yields a high marginal PAR. Accordingly, all investigated methods perform well when the number of D-variants is low, and hence the marginal PAR is high (Figure 3). When the number of D-variants rises, and hence the marginal PAR of each variant drops, the power to identify a disease-group falls (Figure 3). For the weighted-sum method, the effect of the number of D-variants depends on the genetic model. For the recessive models, it is able to identify even large groups of variants, whereas it is more sensitive to the number of D-variants when the heterozygote contributes to disease-risk (Figure 3).
