Assessing genome-wide statistical significance for large p small n problems.

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Assessing genome-wide statistical significance is critical for interpreting genetic study findings.
• Traditional methods may struggle with the high dimensionality of genomic data (large p, small n).

Purpose of the Study:

• To introduce and evaluate a novel resampling approach for determining genome-wide significance thresholds.
• To assess the performance of this method in controlling type I error rates.

Main Methods:

• A new resampling strategy was developed for significance threshold determination.
• Simulations were conducted to evaluate the approach under various genetic data scenarios.

Main Results:

• The proposed resampling approach effectively controls the genome-wide type I error rate.
• Accurate control was observed even in challenging "large p, small n" situations.

Conclusions:

• The novel resampling method provides a robust way to establish genome-wide significance thresholds.
• This approach enhances the reliability of statistical inference in genetic studies, particularly those with complex data structures.