Ancestry-stratified variant classification in monogenic diabetes genes: annotation coverage and differential curation burden
Dario, P.
Show abstract
Variant databases ClinVar and gnomAD are the backbone of clinical variant interpretation, but their population composition is skewed toward European ancestry. Whether this skew creates systematic classification disadvantages for non-European patients with monogenic diabetes has not been examined at the database level. ClinVar variant_summary (GRCh38, April 2026; 4,421,188 variants) was cross-referenced with gnomAD v4.0 genome data for 17 monogenic diabetes genes. Annotation coverage and variant classification rates were computed stratified by genetic ancestry group (AFR, AMR, EAS, SAS, MID, NFE, FIN, ASJ). Of 14,691 gnomAD variants across the 17 genes, only 29.7% had any ClinVar classification (range: 12.7%-61.3% by gene). Among classified variants, non-Finnish European (NFE) variants had the highest variant of uncertain significance (VUS) rate (32.1%) and the lowest benign/likely benign fraction (41.6%), consistent with a large submission volume without functional follow-up. African-ancestry (AFR) variants showed the second-highest VUS rate (29.2%), not statistically distinguishable from NFE after Bonferroni correction, while all other non-European groups had significantly lower rates (all p < 0.001). GCK showed a pattern inversion - non-European VUS rate (18.5%) exceeding European (15.0%) - consistent with progressive reclassification in European populations absent in non-European cohorts. Annotation coverage and VUS divergence were uncorrelated (r = -0.15, p = 0.57). The primary equity problem is a 70% annotation gap combined with a non-European curation deficit, not a simple VUS excess. Ancestry-stratified evaluation of ClinGen Variant Curation Expert Panel (VCEP) criteria performance is warranted across disease domains.
Matching journals
The top 2 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Identification of actionable genetic variants in 4,198 Scottish volunteers from the Viking Genes research cohort and implementation of return of results 95%
- Extracting and calibrating evidence of variant pathogenicity from population biobank data 94%
- Advanced variant classification framework reduces the false positive rate of predicted loss of function (pLoF) variants in population sequencing data 93%
Similar papers in this journal
- Evaluating Genome Sequencing Strategies: Trio, Singleton, and Standard Testing in Rare Disease Diagnosis 93%
- STRchive: a dynamic resource detailing population-level and locus-specific insights at tandem repeat disease loci 93%
- Clustering of predicted loss-of-function variants in genes linked with monogenic disease can explain incomplete penetrance 93%
Similar papers in this journal
- A clinical algorithm to identify people with the glucose-6-phosphate dehydrogenase p.Val68Met variant at risk for diabetes undertreatment 93%
- Genetic Diagnosis of Facioscapulohumeral Muscular Dystrophy Type 1 Using Rare Variant Linkage Analysis and Long Read Genome Sequencing 92%
- Detection and characterisation of copy number variants from exome sequencing in the DDD study 91%
Similar papers in this journal
- Long-read genome sequencing for the diagnosis of neurodevelopmental disorders 93%
- Inclusion of Variants Discovered from Diverse Populations Improves Polygenic Risk Score Transferability 92%
- IMPROVE-DD: Integrating Multiple Phenotype Resources Optimises Variant Evaluation in genetically determined Developmental Disorders 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.