Entropy based genetic association tests and gene-gene interaction tests.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Entropy-based tests, primarily using Shannon entropy, are increasingly used for genetic association studies.
• These methods often show superior statistical power compared to traditional chi-squared tests.
• However, the optimal entropy measure for different genetic models remains an area for exploration.

Purpose of the Study:

• To extend existing entropy-based genetic association tests using Rényi entropy, a generalized form of Shannon entropy.
• To investigate the impact of the order parameter (λ) of Rényi entropy on test power.
• To compare the performance of generalized entropy tests across different genetic models.

Main Methods:

• Application of Rényi entropy, with varying order λ, to single nucleotide polymorphism (SNP) association and gene-gene interaction tests.
• Utilizing genotype and haplotype frequencies to weight events based on their probabilities.
• Conducting simulation studies and real data analyses to evaluate test performance.

Main Results:

• The order 2 Rényi entropy test demonstrated higher power for the dominant genetic model.
• For the multiplicative genetic model, orders 1 (Shannon entropy) and 2 showed comparable power.
• The optimal choice of λ was found to be dependent on the specific genetic model under investigation.

Conclusions:

• Rényi entropy provides a flexible framework for enhancing genetic association tests.
• The selection of the appropriate order λ is crucial for maximizing statistical power.
• Shannon entropy is not universally the most powerful measure; generalized entropy offers a more adaptable approach for genetic association and interaction analyses.