Challenging false discovery rate: a partition test based on p values in human case-control association studies

Area of Science:

• Genetics
• Statistical Genetics
• Bioinformatics

Background:

• Heritable traits are often influenced by multiple genetic variants, each with a small effect.
• Identifying these variants requires robust statistical methods capable of handling complex genetic architectures.

Purpose of the Study:

• To develop a novel statistical test for detecting the association of multiple genetic variants with heritable traits.
• To provide a method that is sensitive to small effect sizes from numerous variants.

Main Methods:

• A partition test was developed to analyze p-values from genome-wide association studies.
• The test identifies sets of smallest p-values that deviate significantly from random distribution.
• This approach differentiates true associations from random occurrences.

Main Results:

• Power calculations indicated the partition test's superiority over single-SNP (SNP-by-SNP) analyses.
• Application to six datasets demonstrated the test's effectiveness in detecting multiple contributing SNPs.
• The partition test yielded more significant results compared to a standard false discovery rate estimation procedure.

Conclusions:

• The partition test effectively identifies multiple functional single nucleotide polymorphisms (SNPs) underlying disease.
• This method can achieve high statistical significance where single-locus tests fail.
• The test also provides an estimate of the number of functional SNPs involved in a trait.