Figure 2

Comparison of the proposed method with MAF partitioning and other statistical tests on population genetics simulations. In the “Rare” phenotype model only rare variants (MAF<1% in haplotype pool) were causal with uniform effect size. “Low Frequency” and “Common” phenotype models had only one low frequency (MAF between 1% and 5%) and one common (MAF>5%) causal SNP respectively. Finally, the “Interaction” scenario models the interaction of rare variants with a common SNP. A minor allele of a rare causal variant had an impact on phenotype if and only if it was present on the same haplotype as a minor allele of a common SNP chosen beforehand.