The posterior probability that SNPs in configuration γ are causal is computed by normalizing over Γ∗ p(γ|y,X)=p∗(γ|y,X)/∑γ∈Γ∗p∗(γ|y,X). We compute the marginal posterior probability that the ℓth SNP is causal, also called single-SNP inclusion probability, by averaging over all evaluated configurations p(γℓ=1|y,X)=∑γ∈Γ∗1(γℓ=1)p(γ|y,X).
