Calculation of exact p-values when SNPs are tested using multiple genetic models.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Multiple comparison methods in genetic association studies often lead to inflated type 1 errors or are too conservative.
• Testing single nucleotide polymorphisms (SNPs) under multiple genetic models (additive, dominant, recessive) is common but complicates statistical analysis.
• Existing methods for multiple testing in genetics struggle with correlated tests and small sample sizes.

Purpose of the Study:

• To develop a novel method for calculating exact p-values for SNPs across different genetic models.
• To address the limitations of current statistical approaches in genetic association studies.
• To enhance the accuracy and power of genetic association analyses.

Main Methods:

• A new statistical method was developed to compute exact p-values for SNPs tested under additive, dominant, and recessive genetic models.
• Simulations were conducted to evaluate the performance of the proposed method.
• The method was applied to analyze a specific polymorphism (eNOS -786T>C) in relation to breast cancer risk.

Main Results:

• Simulations demonstrated that the proposed method effectively controls type 1 errors.
• The new approach resulted in increased statistical power compared to existing methods.
• Application to the eNOS -786T>C polymorphism identified an association with breast cancer risk.

Conclusions:

• The proposed method offers a robust way to analyze genetic data using multiple genetic models.
• It maximizes statistical power while maintaining control over type 1 errors.
• This approach is recommended for genetic association studies employing additive, dominant, and recessive models.