BPHunter

BPHunter detects genome-wide human intronic branchpoint (BP) variants that disrupt pre-messenger RNA splicing and prioritizes candidates for functional and clinical interpretation.

Key Features:

• Integration of Comprehensive BP Data: Integrates a human genome-wide BP database built from existing BP data, RNA sequencing of lariat debranching enzyme DBR1-mutated patients, and machine-learning predictions.
• Characterization of BP Features: Characterizes BP positions and the two upstream nucleotides (BP-2) in major and minor introns, reporting lower variation and higher evolutionary conservation compared to intronic background and comparable to exons.
• Systematic Variant Detection: Systematically detects intronic variants that may disrupt BP recognition and retrospectively identified 40 of 48 known pathogenic BP variants.
• Variant Prioritization Strategy: Provides a prioritization strategy to distinguish benign from potentially deleterious BP variant candidates.
• Practical Utility and Validation: Prospectively identified and experimentally validated a germline heterozygous BP variant in STAT2 linked to critical COVID-19 pneumonia and a somatic intronic deletion in ITPKB in a lymphoma patient.

Scientific Applications:

• Splicing mechanism research: Enables detection and prioritization of BP-disrupting variants to study pre-mRNA splicing mechanisms.
• Genetic disease research and translational genetics: Supports identification of candidate pathogenic BP variants for translational studies, potential therapeutic target identification, and personalized medicine applications.

Methodology:

Integrates comprehensive BP datasets (existing BP annotations, RNA-seq from DBR1-mutated patients) with machine-learning predictions and analyzes sequencing data to identify variants that may disrupt BP recognition, leveraging evolutionary conservation of BP positions.