Prediction of mutant mRNA splice isoforms by information theory-based exon definition

Area of Science:

• Molecular Biology
• Bioinformatics
• Genetics

Background:

• Splicing mutations can lead to diverse mRNA isoforms, complicating molecular phenotypes.
• Accurate prediction of these isoforms is crucial for understanding genetic diseases.

Purpose of the Study:

• To develop and validate a computational method for predicting mRNA isoforms resulting from splicing mutations.
• To assess the accuracy of predicting cryptic and exon-skipping isoforms using splice site information.

Main Methods:

• Utilized information content (R(i)) of splice sites and their distances (gap surprisal) to define exons.
• Calculated total exon information content (R(i),total) and differences (ΔR(i,total)) to predict isoform abundance.
• Employed the Automated Splice Site and Exon Definition Analysis server for in silico predictions.

Main Results:

• The study's predictions of splicing mutations showed high concordance (85.2%) with experimental expression data.
• Differences in total exon information content effectively predicted relative isoform abundance.
• Computational constraints helped eliminate non-conforming and poorly expressed isoforms.

Conclusions:

• In silico exon definition analysis is a powerful tool for predicting mRNA splicing isoforms.
• This approach can streamline the assessment of mutations affecting mRNA splicing.
• The method aids in understanding complex molecular phenotypes arising from splicing alterations.