Abstract / Summary
Population-scale sequencing can quantify genotypes associated with von Willebrand disease (VWD), but these frequencies cannot be translated directly into clinically manifest disease without phenotype and penetrance data. We therefore estimated subtype-specific VWD-associated genotype frequencies in RGC-ME, compared them with gnomAD, and examined which analytical assumptions limit clinical interpretation. Germline VWF variants classified in ClinVar as pathogenic or likely pathogenic (P/LP) on 1 August 2026 were normalized to GRCh38 and matched exactly to both resources. Subtypes were assigned only from unambiguous ClinVar phenotypes, and expected genotype frequencies were calculated from cumulative allele frequencies under dominant or recessive models. ClinVar contained 459 unique sequence-level P/LP alleles; 116 were shared by RGC-ME and gnomAD, 20 occurred only in RGC-ME, 68 only in gnomAD, and 255 in neither extract. The expected RGC-ME genotype frequencies per million individuals were 198, 16, 43, and 224 for types 1, 2A, 2B, and 2M, respectively, and 22.88 and 0.008 for types 2N and 3. Corresponding RGC-ME/gnomAD ratios ranged from 0.753 to 0.911. Shared allele frequencies were strongly concordant (Pearson r = 0.933; Spearman ρ = 0.887), although 238/459 alleles lacked an unambiguous subtype assignment. The cumulative frequency of P/LP VWF alleles was similar across resources, whereas subtype-specific expected genotype frequencies varied with phenotype annotation, the contribution of recurrent alleles, inheritance assumptions, and incomplete variant capture. Population databases can therefore quantify allele burden and model the frequency of VWD-associated genotypes, but clinical registries are required to determine penetrance, genotype–phenotype relationships, and the prevalence of clinically diagnosed VWD.