A new family-based association test via a least-squares method

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Family-based genetic studies are crucial for identifying disease-associated markers.
• Existing likelihood-based methods can be computationally intensive and complex.
• A simplified, robust statistical approach is needed for genetic association analysis.

Purpose of the Study:

• To develop and validate a novel least-squares method for testing genetic marker-phenotype associations using family data.
• To assess the method's performance and comparability with existing approaches.
• To apply the method to identify genetic factors associated with alcoholism.

Main Methods:

• A least-squares approach was developed utilizing phenotype vectors and cross-products within families.
• Covariate adjustment was incorporated into the model.
• The method's asymptotic equivalence to generalized estimating equations with a diagonal working covariance matrix was established.

Main Results:

• The proposed least-squares method is numerically simpler than traditional likelihood-based methods.
• The method demonstrated asymptotic equivalence to generalized estimating equation approaches, addressing prior covariance matrix complexities.
• Application to the Collaborative Study on the Genetics of Alcoholism data revealed a significant association between marker rs1037475 and alcoholism.

Conclusions:

• The novel least-squares method provides an efficient and simpler alternative for family-based genetic association studies.
• This method facilitates covariate adjustment and overcomes limitations of previous approaches.
• The identified association between rs1037475 and alcoholism warrants further investigation in genetic studies of addiction.