Integrating many co-splicing networks to reconstruct splicing regulatory modules

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Alternative splicing significantly expands the proteome but its regulatory mechanisms remain poorly understood.
• Genome-wide studies of coordinated splicing regulation are limited, hindering a comprehensive understanding of this process.

Purpose of the Study:

• To develop a method for identifying splicing modules, defined as sets of cassette exons co-regulated by the same splicing factors, across multiple human RNA-seq datasets.
• To establish a framework for studying genome-wide splicing regulation and deciphering the splicing code.

Main Methods:

• A tensor-based approach was designed to model RNA-seq datasets as co-splicing networks, with nodes representing exons and edges weighted by exon inclusion rate correlations.
• Frequent co-splicing clusters were identified across 38 human RNA-seq datasets.
• Identified clusters were validated against biological knowledge databases to assess their biological significance.

Main Results:

• An atlas of frequent co-splicing clusters, representing potential splicing modules, was generated.
• Validation confirmed that the recurrence of co-splicing clusters across datasets correlates with their biological meaningfulness.
• The integrative approach highlighted the importance of analyzing multiple datasets for robust identification of splicing modules.

Conclusions:

• Co-splicing clusters reveal novel functional groups distinct from co-expression clusters, offering new insights into post-transcriptional regulation.
• Exons can dynamically participate in different functional pathways based on co-splicing partners and experimental conditions.
• Identifying splicing modules is a crucial step towards deciphering the complex splicing code.