Composite statistics for QTL mapping with moderately discordant sibling pairs

Area of Science:

• Human genetics
• Statistical genetics
• Genomic research

Background:

• Extreme discordant sibling-pair (EDSP) designs are theoretically powerful for mapping quantitative trait loci (QTLs).
• Practical application of EDSP designs is limited by the need for large populations to find extremely discordant pairs.
• Current QTL mapping methods like Haseman-Elston regression and variance-components analysis use trait values and identity by descent (IBD) information.

Purpose of the Study:

• To demonstrate the power of moderately discordant sibling pairs for QTL mapping.
• To develop a more practical balance between phenotyping and genotyping efforts in genetic studies.
• To propose and validate a composite statistic for enhanced linkage analysis.

Main Methods:

• Developing a new statistical result for analyzing moderately discordant pairs.
• Combining identity by descent (IBD) sharing statistics with trait value given IBD statistics.
• Testing a novel composite statistic using simulation studies and a real-world dataset.

Main Results:

• Moderately discordant pairs provide substantial power for QTL mapping.
• The proposed composite statistic minimally increases power for extremely discordant pairs.
• The composite statistic significantly boosts power for moderately discordant pairs, enhancing practicality.

Conclusions:

• Moderately discordant sibling pairs are a practical alternative for QTL mapping.
• Combining different statistical approaches (IBD sharing and trait values) enhances linkage analysis power.
• The developed composite statistic is computationally feasible with existing software.