ncVarDB

ncVarDB curates evidence-linked annotations of pathogenic and benign human non-coding variants to support interpretation of regulatory, splicing, and non-coding RNA effects on gene function.

Key Features:

• Curated non-coding variants: ncVarDB contains 721 characterized human non-coding variants with phenotypic consequences supported by published evidence and verified by two independent curators.
• Benign controls: The resource includes 7,228 covariate-matched benign controls (population frequency > 5%) selected from dbSNP151, matched by annotation type (UTR, intronic, intergenic) and variant type (substitution or indel) and with minimized linkage disequilibrium to pathogenic variants.
• Curation workflow: Entries were compiled through a structured curation workflow with dual independent verification of pathogenicity-related annotations to reduce annotation error.

Scientific Applications:

• Method development and benchmarking: Dataset supports development, training, and evaluation of computational methods that predict functional impact of non-coding variants.
• Regulatory and disease variant analyses: Enables analyses of regulatory variant contributions to disease, including effects on gene regulation, splicing, and non-coding RNA function.

Methodology:

ncVarDB was constructed using a structured curation workflow with dual independent verification of pathogenicity-related annotations.