RegSNPs-intron: a computational framework for predicting pathogenic impact of intronic single nucleotide variants

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Intronic single nucleotide variants (SNVs) are understudied for their disease-causing potential.
• Systematic investigation of intronic SNVs (iSNVs) is crucial for understanding genetic disorders.
• Existing methods lack comprehensive evaluation of iSNVs' impact on molecular mechanisms.

Purpose of the Study:

• To develop and validate a computational tool for predicting the pathogenicity of iSNVs.
• To integrate diverse biological features for accurate iSNV impact assessment.
• To establish a high-throughput experimental method for validating iSNV effects on RNA splicing.

Main Methods:

• Development of the RegSNPs-intron algorithm using a random forest classifier.
• Integration of RNA splicing, protein structure, and evolutionary conservation features.
• Validation using a high-throughput functional reporter assay, ASSET-seq (ASsay for Splicing using ExonTrap and sequencing).

Main Results:

• RegSNPs-intron demonstrated excellent performance in evaluating pathogenic impacts of iSNVs.
• ASSET-seq successfully assessed the functional consequences of iSNV predictions on splicing.
• The combined approach effectively prioritizes iSNVs for disease pathogenesis studies.

Conclusions:

• RegSNPs-intron provides a robust method for assessing iSNV pathogenicity.
• ASSET-seq offers a scalable platform for functional validation of splicing alterations.
• The integrated strategy enhances the identification of disease-associated intronic variants.