Genetic association analysis using data from triads and unrelated subjects

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Selecting appropriate control samples is crucial for accurate genetic association studies.
• Unrelated controls risk spurious associations due to population stratification.
• Parental controls (triads) are robust to stratification but difficult to collect.

Purpose of the Study:

• To develop a unified statistical framework for analyzing combined genetic data from both triad and case-control studies.
• To improve the power and accuracy of genetic association analyses by leveraging diverse control sample types.

Main Methods:

• A likelihood-based approach was developed to integrate data from triads and unrelated controls.
• The method allows for joint analysis of case-control and family-based study designs.
• Simulations were used to evaluate the performance of the proposed method.

Main Results:

• The combined analysis approach demonstrated greater statistical power compared to analyzing each sample type separately.
• The method provides flexible modeling for allele effects and accommodates missing parental data.
• The approach was successfully illustrated using SNP data from a psoriasis candidate-gene study.

Conclusions:

• The proposed likelihood-based method effectively combines information from diverse control samples in genetic association studies.
• This unified approach offers a more powerful and flexible alternative to analyzing separate control groups.
• The method has practical implications for candidate-gene studies and potentially genome-wide association studies.