Genome-wide detection of human variants that disrupt intronic branchpoints

Area of Science:

• Molecular Biology
• Genetics
• Bioinformatics

Background:

• Pre-messenger RNA splicing relies on recognizing intronic branchpoint (BP) sequences by spliceosome elements.
• Disruptions in BP sequences by rare variants can lead to disease, but efficient detection in sequencing data was lacking.
• Forty-eight pathogenic variants affecting BP have been identified, highlighting the clinical significance of BP integrity.

Purpose of the Study:

• To develop a genome-wide computational approach for efficiently detecting intronic variants that disrupt branchpoint recognition.
• To establish a comprehensive human genome-wide BP database.
• To identify novel disease-causing variants associated with splicing defects.

Main Methods:

• Integrated existing BP data with new data from DBR1-mutated patients and machine-learning predictions to create a genome-wide BP database.
• Characterized BP and BP-2 features, assessing variation rates and evolutionary conservation.
• Developed and applied BPHunter, a computational tool, to retrospectively and prospectively identify BP variants in sequencing data.

Main Results:

• BPHunter identified 40 of 48 known pathogenic BP variants, providing a strategy for prioritizing candidates.
• BP and BP-2 positions show low variation and high conservation, comparable to exonic regions.
• Prospectively identified novel pathogenic variants in STAT2 (COVID-19) and ITPKB (lymphoma), validated experimentally.

Conclusions:

• BPHunter is an effective genome-wide computational tool for systematically detecting intronic variants disrupting BP recognition.
• The study provides a valuable resource and method for identifying genetic variants underlying splicing-related diseases.
• BPHunter has demonstrated practical utility in identifying novel disease-causing variants, aiding clinical diagnostics and research.