A weighted U statistic for association analyses considering genetic heterogeneity.

Area of Science:

• Genetics
• Statistical genetics
• Complex diseases

Background:

• Complex diseases can share symptoms but have different genetic causes.
• Genetic studies often assume uniform genetic effects, potentially reducing power.
• Genetic heterogeneity in complex diseases is an under-addressed issue in current research.

Purpose of the Study:

• To propose a novel statistical method, heterogeneity-weighted U (HWU), to address genetic heterogeneity in association analyses.
• To develop a computationally efficient method applicable to high-dimensional genetic data and various phenotypes.
• To improve the power of genetic studies for complex diseases with diverse etiological pathways.

Main Methods:

• Developed the heterogeneity-weighted U (HWU) method for genetic association studies.
• Validated the HWU method through simulations, assessing its performance under heterogeneous genetic etiology and robustness to model assumptions.
• Applied the HWU method to a genome-wide analysis of nicotine dependence using the Study of Addiction: Genetics and Environments dataset.

Main Results:

• Simulations demonstrated HWU's advantage in detecting genetic effects when disease etiology is heterogeneous.
• HWU showed robustness across different phenotype distribution assumptions.
• Genome-wide analysis identified heterogeneous effects of CYP3A5 and IKBKB genes on nicotine dependence.

Conclusions:

• The proposed HWU method effectively accounts for genetic heterogeneity in complex diseases.
• HWU enhances the power of genetic association studies and can identify novel disease-related genes.
• This approach advances the understanding of genetic underpinnings for conditions like nicotine dependence.