Adjusted p-value Thresholds for Genome Wide Association Studies
Author Information
Author(s): Duggal Priya, Gillanders Elizabeth M, Holmes Taura N, Bailey-Wilson Joan E
Primary Institution: National Human Genome Research Institute, National Institutes of Health
Hypothesis
What is the appropriate method to control the family-wide Type 1 Error in genetic association studies?
Conclusion
The study provides adjusted p-value thresholds that account for the interdependence of SNPs, improving the accuracy of genome-wide association studies.
Supporting Evidence
- The study reduced the number of effective tests from 500,000 to 67,000 for the Affymetrix SNP panel.
- Adjusted p-value thresholds were established for both suggestive and significant associations.
- The proposed method accounts for the interdependence of SNPs, improving the accuracy of significance testing.
Takeaway
This study helps scientists figure out how to better analyze genetic data by using smarter ways to set significance levels, so they don't mistakenly think something is important when it's not.
Methodology
The study estimated the effective number of independent SNPs by analyzing genotype data from the International HapMap project and applying a modified Bonferroni correction.
Potential Biases
The study acknowledges the potential for overcorrection in SNPs that are not truly independent.
Limitations
The proposed thresholds are population-specific and may vary across different genetic backgrounds.
Participant Demographics
The study utilized data from the CEPH Utah (CEU) and Yoruba (YRI) populations.
Statistical Information
P-Value
≈10-5, 10-7, 10-8
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website