Correcting for cryptic relatedness by a regression-based genomic control method.

Area of Science:

• Population Genetics
• Statistical Genetics
• Genomic Association Studies

Background:

• The genomic control (GC) method corrects for cryptic relatedness in population-based association studies.
• Traditional GC requires matching null markers to candidate markers by allele frequencies, limiting its use.
• Allele frequency errors can exacerbate bias and variance inflation, complicating GC correction.

Purpose of the Study:

• To propose a generalized regression-based genomic control (GC) method.
• To overcome the limitation of allele frequency matching in traditional GC methods.
• To improve the applicability and robustness of GC for association studies.

Main Methods:

• Introduced a regression-based GC method utilizing null markers without strict allele frequency matching.
• Employed a regression approach to adjust for variations in null marker allele frequencies.
• Validated the method's performance through simulations.

Main Results:

• The proposed regression-based GC method effectively adjusts for allele frequency variations.
• The method demonstrates applicability beyond the Cochran-Armitage additive trend test.
• Simulation results confirm the method's effectiveness in controlling type I error.

Conclusions:

• The novel regression-based GC method expands the utility of genomic control in association studies.
• It is readily applicable to various statistical tests, including Cochran-Armitage trend tests and Pearson's chi-square test.
• The method successfully controls type I error rates in the presence of population substructure.