Family-based association tests for different family structures using pooled DNA

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• DNA pooling is a cost-effective method for genomewide association studies (GWAS) to identify disease genes.
• Existing family-based association study methods often assume uniform family structures, which is not typical in real genetic studies.
• The informativeness of different family structures for genetic association analysis is dependent on the unknown disease model.

Purpose of the Study:

• To develop and investigate statistical methods for combining information from diverse family types in genetic association studies.
• To propose a general strategy for incorporating different family structures by assigning optimal weights in association tests.
• To account for the impact of measurement errors on the required sample size in DNA pooling studies.

Main Methods:

• Proposed a general strategy to incorporate diverse family types by assigning optimal weights in association tests.
• Incorporated measurement errors into the analysis of DNA pooling data.
• Evaluated the proposed weighting scheme under various disease models and measurement error rates.

Main Results:

• The proposed weighting scheme significantly reduces the required sample size compared to previous approaches (Risch & Teng, 1998).
• Measurement errors can have a substantial impact on the required sample size, particularly when error rates are not negligible.
• The effectiveness of the weighting scheme is demonstrated across different disease models.

Conclusions:

• The developed weighting strategy offers a more efficient approach to DNA pooling in family-based genetic association studies by leveraging diverse family structures.
• Accounting for measurement errors is crucial for accurate sample size estimation in DNA pooling studies.
• This method provides a robust framework for identifying disease genes using cost-effective DNA pooling strategies with heterogeneous family data.