Mixed-effects logistic approach for association following linkage scan for complex disorders

Area of Science:

• Genetics
• Statistical genetics
• Complex disease genetics

Background:

• Association studies are crucial for dissecting complex disorders but face challenges with family structures.
• Traditional methods using only independent cases/controls or single affected individuals per family have limitations in power and efficiency.
• Using all affected family members with unrelated controls can lead to false positives due to relatedness.

Purpose of the Study:

• To propose a new statistical approach for genetic association studies that incorporates family members and unrelated controls.
• To develop an efficient algorithm for simulating genetic and environmental factors within families.
• To evaluate the performance of the proposed method in terms of error control and statistical power.

Main Methods:

• Development of a mixed-model logistic regression framework for association analysis.
• Creation of a simulation algorithm to model genotypes, environmental risk factors, and linkage disequilibrium (LD) in families.
• Extensive simulation studies to assess type-I error probability and statistical power.

Main Results:

• The proposed mixed-model approach effectively controls type-I error probability.
• The method demonstrates higher statistical power compared to traditional family-based designs (e.g., TDT) and other likelihood-based methods, especially when environmental factors are involved.
• The approach allows for the inclusion of covariates like age and smoking status and can be extended to complex pedigrees.

Conclusions:

• The novel mixed-model logistic regression approach offers a powerful and flexible tool for genetic association studies of complex disorders.
• This method overcomes limitations of existing designs by leveraging family data efficiently while controlling for relatedness and incorporating environmental factors.
• It provides a robust framework for identifying genetic variants contributing to complex diseases.