A hierarchical model of evidence for establishing the pathogenicity of a sequence variant is shown. Identification of a novel variant in an affected individual provides relatively weak evidence because it does not address the possibility that the variant is benign. Somewhat stronger evidence comes from prediction of a variant's impact on gene function, for example, changes in protein-coding sequence or known regulatory regions or disruption of an evolutionarily conserved domain. Association studies of phenotype and candidate mutations in cases and controls and linkage studies of candidate mutations within families offer further support for a variant's clinical significance. The most conclusive evidence of pathogenicity is provided by functional studies of sequence variants: for example, introduction of a candidate mutation in vivo recapitulates the disease phenotype or affects protein function in an in vitro assay. As demonstrated by Bellissimo et al, undergirding all of these studies are investigations of ethnically diverse normal populations, which help to illuminate normal genetic variation.⁵