MutPred Splice: machine learning-based prediction of exonic variants that disrupt splicing

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Pre-mRNA splicing is crucial for gene expression.
• Mutations in coding regions can alter splicing patterns.
• Accurate identification of splicing-disrupting mutations is vital for disease understanding.

Purpose of the Study:

• To develop a novel machine-learning approach for identifying coding region substitutions that disrupt pre-mRNA splicing.
• To estimate the frequency of splicing disruption in inherited diseases and cancer.

Main Methods:

• Development of MutPred Splice, a machine-learning tool.
• Application of MutPred Splice to analyze human disease-causing exonic mutations.
• Analysis of mutation mechanisms affecting splicing.

Main Results:

• MutPred Splice successfully identifies mutations disrupting pre-mRNA splicing.
• 16% of inherited disease mutations and 10-14% of cancer somatic mutations may disrupt splicing.
• Inherited diseases primarily involve splice site loss, while cancer involves exon skipping via splicing regulatory elements.

Conclusions:

• MutPred Splice is a valuable tool for identifying splicing-disrupting mutations.
• Splicing defects are a significant factor in both inherited diseases and cancer.
• Distinct molecular mechanisms underlie splicing disruption in different disease contexts.