Evaluation of gene-based association tests for analyzing rare variants using Genetic Analysis Workshop 18 data

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Rare variants play a crucial role in complex diseases.
• Evaluating association tests for rare variants requires robust statistical methods.
• Gene annotation and binning strategies can influence rare variant analysis outcomes.

Purpose of the Study:

• To assess the performance of novel pooled association tests for rare variants.
• To investigate the impact of gene annotation and binning strategies on rare variant analysis.
• To evaluate these methods using real and simulated data from the Genetic Analysis Workshop 18.

Main Methods:

• Utilized Genetic Analysis Workshop 18 data including sequence genotypes, hypertension status, and simulated systolic/diastolic blood pressure (SBP/DBP).
• Applied various pooled association tests, gene annotation methods, and binning strategies.
• Analyzed rare variants on chromosome 3 in 103 unrelated individuals.

Main Results:

• No significant results were found for real hypertension data across binning strategies, though protein-damaging variant grouping showed minor p-value deviations.
• Simulated data analysis revealed low power for most causal genes, except for MAP4, which showed good power for both SBP and DBP.
• Fisher's method proved robust, enhancing power over individual tests and outperforming other robust tests like SKAT-O.

Conclusions:

• Fisher's method is a reliable and powerful approach for rare variant association studies, especially for quantitative traits like blood pressure.
• The choice of gene annotation and binning strategy can impact the detection of rare variant associations.
• MAP4 represents a potential genetic locus for blood pressure regulation, warranting further investigation.