JUM is a computational method for comprehensive annotation-free analysis of alternative pre-mRNA splicing patterns

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Alternative pre-mRNA splicing (AS) is vital for gene regulation and proteome diversity in metazoans.
• Current AS analysis methods require pre-existing transcriptome annotations, limiting discovery in poorly annotated genomes.

Purpose of the Study:

• To develop a novel, annotation-independent computational method for comprehensive AS analysis.
• To accurately identify, analyze, and quantify global AS profiles, including novel events and intron retention.

Main Methods:

• Developed the junction usage model (JUM), a bottom-up approach for AS analysis.
• JUM quantifies AS patterns without relying on annotated transcript libraries.
• Stringently classifies intron retention (IR) to minimize false positives.

Main Results:

• JUM identified approximately twice the number of novel AS events compared to existing methods in diverse RNA samples.
• Demonstrated a 1.2 to 4.8 times higher true positive rate for AS detection.
• Significantly reduced false positive rates for intron retention detection.

Conclusions:

• JUM provides a robust, annotation-free framework for analyzing complex metazoan transcriptomes.
• Enables superior accuracy in detecting and quantifying diverse alternative splicing patterns.
• Facilitates AS research in organisms with limited genomic annotation.