A Bayesian method to estimate variant-induced disease penetrance

Area of Science:

• Genomic Medicine
• Computational Biology
• Cardiovascular Genetics

Background:

• Genomic medicine faces challenges in assessing disease risk from rare or novel genetic variants.
• Large-scale phenotyping and DNA sequencing data offer opportunities to refine genetic risk assessments.
• Accurate estimation of variant penetrance is crucial for personalized disease risk prediction.

Purpose of the Study:

• To propose and validate a computational framework for estimating disease probability (penetrance) based on genetic variant features.
• To apply this framework to the SCN5A gene and Brugada syndrome, a well-characterized disease-gene pair.
• To identify key variant attributes predictive of Brugada syndrome penetrance.

Main Methods:

• Developed a pattern mixture algorithm using a Bayesian Beta-Binomial model.
• Reviewed 756 publications to gather data on SCN5A variants and Brugada syndrome.
• Incorporated variant-specific attributes (function, structural context, sequence conservation) and heterozygote observations into the model.

Main Results:

• Generated SCN5A variant penetrance probabilities for Brugada syndrome.
• Identified variant functional perturbation and structural context as the most predictive features for Brugada syndrome penetrance.
• The framework provides a quantitative approach to disease risk assessment for genetic variants.

Conclusions:

• The proposed framework offers a robust method for estimating genetic variant penetrance.
• Variant functional and structural characteristics are key determinants of disease risk in Brugada syndrome.
• This approach can be extended to other disease-gene pairs to improve genomic medicine risk prediction.