Joint Analysis of Multiple Phenotypes in Association Studies based on Cross-Validation Prediction Error.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Joint analysis of multiple phenotypes in genome-wide association studies (GWAS) offers greater statistical power for identifying genetic variants linked to complex diseases compared to single-phenotype analyses.
• Existing methods like O'Brien's, TATES, MANOVA, and MultiPhen show inconsistent performance across various simulation scenarios.
• A key challenge is developing a joint analysis method that maintains robust performance across diverse situations.

Purpose of the Study:

• To introduce a novel statistical method, Multiple Phenotypes-Prediction Error (MultP-PE), for testing associations between genetic variants and multiple phenotypes.
• To evaluate the performance of MultP-PE in terms of type I error rates and statistical power through extensive simulations.
• To compare MultP-PE against established joint analysis methods.

Main Methods:

• Development of the MultP-PE statistical method, which utilizes cross-validation prediction error.
• Conducting extensive simulation studies to assess type I error rates and statistical power.
• Comparative analysis of MultP-PE against O'Brien's method, TATES, MANOVA, and MultiPhen.

Main Results:

• MultP-PE effectively controls type I error rates across all simulated scenarios.
• MultP-PE demonstrates consistently higher statistical power compared to the evaluated existing methods in all tested scenarios.
• The proposed method shows robust performance irrespective of the simulation conditions.

Conclusions:

• MultP-PE is a reliable and powerful method for joint analysis of multiple phenotypes in GWAS.
• The method's consistent high power and accurate error control make it suitable for various genetic association studies.
• MultP-PE is recommended for its superior performance in identifying genetic variants associated with multiple complex traits.