The power to detect linkage disequilibrium with quantitative traits in selected samples

Area of Science:

• Genetics
• Statistical Genetics
• Human Genetics

Background:

• Phenotypically extreme nuclear families are collected for linkage detection.
• Dense marker maps will enable both linkage and association studies.
• Linkage optimal selection strategies may not be ideal for association studies.

Purpose of the Study:

• To examine the power of different phenotypic selection strategies for detecting linkage disequilibrium in quantitative trait families.
• To compare the effectiveness of six common selection strategies.

Main Methods:

• Evaluated six selection strategies: single selection (2 designs), affected sib pairs, concordant/discordant pairs, and extreme-concordant/-discordant designs.
• Analyzed the impact of allele frequencies on power within each design.
• Assessed power for detecting linkage disequilibrium for quantitative traits.

Main Results:

• Extreme proband selection (one extreme proband) offers power comparable to discordant sib pairs but requires fewer initial families.
• Common trait alleles generally increase power, but marker- and trait-allele frequency similarity is crucial.
• Single selection designs are powerful only for rare marker and quantitative trait loci (QTL) alleles.
• Discordant pairs and extreme-proband designs offer power across a wide range of allele frequency differences.

Conclusions:

• Extreme proband selection provides the best balance of power, robustness, and ascertainment simplicity for family-based association analysis.
• Selection strategy choice significantly influences the success of detecting linkage disequilibrium for quantitative traits.