GABRA1 Missense Variants for Precision Diagnosis of Epilepsy

Wen-Hui Liu curated a dataset of 61 GABRA1 missense variants, including 30 pathogenic/likely pathogenic variants from patients and 31 benign/likely benign controls from gnomAD. The dataset was used to analyze molecular subregional effects and evaluate the performance of 34 pathogenicity prediction algorithms. The work was published on figshare in May 2026.

Use Cases

• Benchmarking pathogenicity prediction algorithms based on the evaluation of 34 computational tools.
• Analyzing molecular subregional effects on phenotype based on variant location mapping.
• Training ensemble or deep learning models for variant classification based on the curated dataset of 61 variants.
• Supporting clinical genetic diagnosis of GABRA1-associated disorders based on the performance analysis of predictive tools.

Strengths

• Dataset includes 61 specifically curated variants with known pathogenicity labels.
• Performance metrics for 34 algorithms are systematically analyzed, including accuracy, AUC, and MCC.
• The study identifies domain-dependent phenotypic effects, providing biological context.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Dataset size is 43.2 KB, indicating a very limited scope.

Provenance

Source

figshare, author Wen-Hui Liu.

Collection Method

Curated from patient data and the gnomAD database.

Freshness

Last updated 2026-05-18 22:02:40; freshness should be verified.