Locus-specific heritability estimation via the bootstrap in linkage scans for quantitative trait loci

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Quantitative trait loci (QTL) linkage analysis heritability estimates are biased when using original data for both linkage localization and effect estimation.
• Genome-wide significance levels exacerbate positive bias in heritability estimates.
• Bias in heritability estimates impacts the accuracy of genetic studies.

Purpose of the Study:

• To propose multi-locus bootstrap resampling estimators for bias reduction in QTL linkage analysis.
• To address bias in heritability estimates when multiple linkage peaks are of interest.
• To improve the accuracy of genetic effect estimations in genome-wide studies.

Main Methods:

• Bootstrap estimates were generated using repeated sample splitting of the original dataset.
• Simulation studies were conducted with nuclear families having 0 to 5 QTLs.
• Methods were applied to a genome-wide blood pressure phenotype analysis in Framingham Heart Study (FHS) pedigrees.

Main Results:

• Bootstrap estimates demonstrated reduced bias and smaller mean squared error compared to naive estimates in simulations.
• In FHS pedigrees, bootstrap heritability estimates were up to 70% smaller than original sample estimates.
• Bias reduction was observed in heritability estimations.

Conclusions:

• Bias-reduced heritability estimates are crucial for accurate genetic effect assessment.
• Using bias-reduced estimates for sample size calculations increases the likelihood of successful replication studies.
• The proposed bootstrap method enhances the reliability of heritability estimates in genetic research.