Multiple subsampling of dense SNP data localizes disease genes with increased precision.

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Linkage studies use dense single nucleotide polymorphism (SNP) data to localize trait loci.
• Correlations between nearby SNPs can distort trait location estimates.
• Existing methods include ignoring correlation, approximating correlation, or using uncorrelated SNP subsets.

Purpose of the Study:

• To develop and test an efficient estimator for trait location using dense SNP data.
• To address the challenge of SNP correlation in linkage analysis.
• To improve the precision of candidate gene region identification.

Main Methods:

• A novel estimator averaging location estimates from random SNP subsamples.
• Ensuring approximate uncorrelation within subsamples using pairwise correlation estimates.
• Employing nonparametric bootstrap for high-resolution confidence intervals (candidate gene regions).

Main Results:

• Existing methods can produce biased or inefficient trait localization estimates.
• The proposed estimator outperforms existing methods in terms of mean squared error.
• A 47.5% reduction in candidate gene region length was achieved using the new method on hypertension family data.

Conclusions:

• The developed method provides a valuable tool for high-resolution candidate gene region construction.
• This approach can aid in efficiently targeting regions for subsequent sequencing projects.
• Improved localization accuracy facilitates gene discovery in genetic studies.