The mid p-value in exact tests for Hardy-Weinberg equilibrium

Area of Science:

• Population genetics
• Statistical genetics

Background:

• Exact tests for Hardy-Weinberg equilibrium are crucial in genetic association studies.
• The standard p-value in these tests can be overly conservative, particularly with small minor allele frequencies.

Purpose of the Study:

• To evaluate the mid p-value as a superior alternative to the standard p-value in exact tests for Hardy-Weinberg equilibrium.
• To assess the performance of the mid p-value regarding type 1 error rate and statistical power.

Main Methods:

• Calculated type 1 error rates and power for exact tests under various conditions (sample size, significance level, minor allele count, deviation from equilibrium).
• Compared three p-value types: standard two-sided, doubled one-sided, and mid p-value.
• Utilized HapMap datasets and a colon cancer dataset to analyze practical implications.

Main Results:

• The mid p-value demonstrated type 1 error rates closer to the nominal level and superior statistical power compared to the standard p-value.
• Differences were more pronounced for non-significant results; the mid p-value identified more significant markers in empirical datasets.
• The equilibrium null distribution was found untenable for both analyzed databases.

Conclusions:

• The standard exact p-value is overly conservative, especially for low minor allele frequencies.
• The mid p-value improves upon this by bringing rejection rates closer to the nominal level, with a slight risk of exceeding it.
• The mid p-value is recommended for genetic association studies for more accurate and powerful equilibrium testing.