Variant classification of hereditary cancer genes is affected by genomic underrepresentation of admixed populations

Variant classification in genetic testing often culminates in "uncertain" calls, known as variants of uncertain significance (VUS), which remain a major clinical challenge. Among the established criteria for variant classification, population allele frequency (AF) is fundamental, yet under-representation of non-European groups hinders accurate interpretation. In this study, we evaluated the impact of population-specific AF on the reclassification of VUS and conflicting variants in hereditary cancer genes. From ClinVar, we curated 487 variants present in both the Brazilian cohort ABraOM and gnomAD v4.1 databases. Comparative population analysis revealed that 43% of shared variants showed significantly different AFs (q ≤ 0.01), with 113 (23%) exhibiting large effect sizes (OR ≥ 4), including 39 VUS. Among these, 20 VUS had higher AF in the Brazilian cohort and exceeded benignity thresholds (BS1), while remaining rare in other populations. Functional prediction tools such as REVEL and CADD failed to distinguish these variants from globally rare VUS. Integrating Brazilian‑specific AF with ClinGen VCEP rules downgraded 15% (3/20) of candidate VUS and resolved five conflicting calls. These findings argue for routine incorporation of regional reference datasets in diagnostic curation to reduce uncertainty and avoid inappropriate clinical management in diverse populations.