Detection of aberrant splicing events in RNA-seq data using FRASER

Area of Science:

• Genomics
• Computational Biology
• Rare Diseases

Background:

• Aberrant splicing is a significant cause of rare genetic diseases.
• Predicting aberrant splicing solely from genome sequences is often inconclusive.
• RNA sequencing offers a complementary approach for detecting aberrant splicing.

Purpose of the Study:

• To develop FRASER, an algorithm for detecting aberrant splicing events from RNA sequencing data.
• To improve the identification of splicing alterations in rare disease diagnostics.

Main Methods:

• FRASER algorithm development for RNA sequencing data analysis.
• Incorporation of alternative splicing and intron retention event detection.
• Implementation of latent confounder control and count distribution-based analysis.
• Application of multiple testing correction for robust statistical inference.

Main Results:

• FRASER detects both alternative splicing and intron retention, doubling the number of aberrant events identified.
• A pathogenic intron retention event in MCOLN1 causing mucolipidosis was identified.
• FRASER significantly reduces false positive calls compared to z-score cutoffs.
• Reprioritization of a pathogenic aberrant exon truncation in TAZ using FRASER.

Conclusions:

• FRASER enhances the detection of aberrant splicing events, including intron retention.
• The algorithm improves the diagnostic yield for rare diseases by identifying pathogenic splicing alterations.
• FRASER offers a robust, user-friendly, and freely available tool for rare disease diagnostics.