ABSTRACT
Background
Variant databases ClinVar and gnomAD underpin clinical variant interpretation, but their composition is skewed toward European ancestry. Whether this produces systematic disadvantages for non-European patients with monogenic diabetes has not been examined at the database level with mutually exclusive ancestry groups.
Objective
To quantify ClinVar annotation coverage and ancestry-stratified classification outcomes across monogenic diabetes genes, using mutually exclusive population-private variant groups.
Methods
gnomAD v4.0 (genomes and exomes; 807,162 individuals) was cross-referenced with the ClinVar GRCh38 record (accessed June 29, 2026) for 16 monogenic diabetes (maturity-onset diabetes of the young [MODY]) genes, and variants were classified as population-private to European or non-European ancestry from ancestry-specific allele counts.
Results
Across 54,865 gnomAD variants, 86.7% carried no ClinVar classification. The annotation gap was near-universal rather than ancestry-specific (unannotated fraction 89.4% European-private, 87.5% non-European-private). Among classified population-private variants, however, the pathogenic or likely pathogenic rate was 19.8% for European-private but only 8.6% for non-European-private variants (Fisher exact OR 2.6, 95% CI 2.1–3.3, p = 1 × 10− 1 8), with a higher uncertain-significance rate in non-European-private variants (52.1% vs. 45.8%).
Conclusion
The dominant problem is therefore an annotation deficit affecting all ancestries; the ancestry inequity the data support is one of actionability, not of raw uncertain-significance rates. Closing it requires ClinVar submissions and population-specific sequencing from underrepresented populations.