Analytic P-value calculation for the higher criticism test in finite d problems.

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• The higher criticism test is effective for detecting sparse effects in genetic association studies.
• Accurate p-value calculation for higher criticism typically requires a large number of genetic markers (d), which is often not feasible for genes or pathways.
• Existing methods rely on asymptotic distributions or simulations, limiting their accuracy for smaller d.

Purpose of the Study:

• To develop an analytic method for accurate p-value computation of the higher criticism test for finite d problems.
• To provide an exact p-value calculation method that does not rely on asymptotic assumptions or simulations.
• To offer a computationally advantageous method for genetic pathway analysis with a limited number of markers.

Main Methods:

• Proposed an analytic method to compute exact p-values for the higher criticism test.
• The method is exact for arbitrary d when test statistics are normally distributed.
• Applied the method to a case-control genome-wide association study of lung cancer.

Main Results:

• The analytic method accurately computes p-values for the higher criticism test with finite d.
• The method is computationally advantageous, especially when d is not large.
• Demonstrated the method's utility in a lung cancer GWAS.

Conclusions:

• The proposed analytic method provides an exact and efficient way to calculate p-values for higher criticism tests in genetic studies with limited markers.
• This method overcomes limitations of asymptotic approximations and simulations, offering improved accuracy and computational efficiency.
• The approach is particularly valuable for analyzing genetic pathways and individual genes in genome-wide association studies.