Four genetic models are investigated (see Figure 1). For the Recessive, Additive and Dominant models the disease-related variants act independently, whereas for the Recessive-Set model the outcome of a mutation at one variant depends on the presence of a mutation at another variant (see Figure 1). We do not sample Dominant-Set or Additive-Set models, since in these models the heterozygote predisposes for disease, and hence they perform like the Dominant and Additive models respectively. We sample the variants independently for simplicity and because rare variants are not expected to be in high LD with the surrounding variants [31],[32].
