Mixture modelling as an exploratory framework for genotype-trait associations.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Classical mixed-effects models have limitations in analyzing genotype-trait associations.
• Existing methods often face challenges with degrees-of-freedom and restrictive distributional assumptions.

Purpose of the Study:

• To propose a flexible mixture modeling framework for identifying and exploring genotype-trait associations.
• To extend classical mixed-effects models by incorporating Gaussian mixture distributions for random genotype effects.

Main Methods:

• Developed a mixture modeling framework integrating Gaussian mixture distributions for random genotype effects.
• Extended classical mixed-effects modeling for genotype-trait association analysis.
• Applied the framework to a study of antiretroviral-associated dyslipidaemia in human immunodeficiency virus infection.

Main Results:

• The proposed framework addresses the degrees-of-freedom challenge inherent in fixed-effects analysis of covariance.
• It relaxes the restrictive single normal distribution assumption of classical mixed-effects models.
• The framework provides an exploratory tool for uncovering structure across multiple genetic loci.

Conclusions:

• The mixture modeling framework offers a robust and flexible approach for genotype-trait association studies.
• It enhances the discovery of genetic architecture underlying complex traits.
• The method shows promise for applications in human immunodeficiency virus research and beyond.