Comparison of Type I Error for Multiple Test Corrections in SNP Studies
Author Information
Author(s): Kristin K Nicodemus, Wenlei Liu, Gary A Chase, Ya-Yu Tsai, M Daniele Fallin
Primary Institution: Johns Hopkins Bloomberg School of Public Health
Hypothesis
Can different methods for correcting type I error rates in SNP studies provide more accurate results?
Conclusion
Using linkage disequilibrium block-based corrections can prevent type I error inflation in SNP studies without being overly conservative.
Supporting Evidence
- Permutation testing is time-consuming but was previously the gold standard for significance levels.
- Bonferroni correction is not suitable due to correlations between tests in SNP settings.
- The study found that the blocking algorithm corrections are closest to the desired type I error under moderate LD.
Takeaway
This study looks at how to avoid mistakes when testing many genes at once. It finds a better way to check if the results are real.
Methodology
The study simulated parent-child trio data under different linkage disequilibrium conditions and compared various correction methods for type I error rates.
Limitations
The study did not consider low LD conditions, which may affect the applicability of the findings.
Participant Demographics
The study used simulated data from the Collaborative Study on the Genetics of Alcoholism.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website