SpliceVarDB

SpliceVarDB consolidates over 50,000 experimentally validated human splicing variants from over 8,000 human genes to improve variant interpretation and support in silico splicing predictor development.

Key Features:

• Spliceogenicity Classification: Variants are classified as "splice-altering" (~25%), "not splice-altering" (~25%), or "low-frequency splice-altering" (~50%) based on the strength of supporting evidence.
• Non-Canonical Variant Coverage: 55% of splice-altering variants are outside canonical splice sites, including 5.6% that are deep intronic.
• Harmonized Validation Data: Data from over 500 published sources are standardized and consolidated across diverse experimental protocols using an established spliceogenicity scale.
• In Silico Predictions: The database incorporates in silico splicing predictions alongside experimental validation metrics.

Scientific Applications:

• Variant Curation: Splice-altering variants support the diagnostic variant curation pathway in clinical settings.
• In Silico Predictor Development: High-quality validated data enables development of more accurate computational splicing predictors.
• Diagnostic Efficiency: Consolidated data reduces duplication of RNA-based validation efforts in diagnostic laboratories.

Methodology:

Over 500 published data sources were evaluated, and a spliceogenicity scale was established to standardize, harmonize, and consolidate variant validation data generated by a range of experimental protocols.