Importance Sampling Method for Multiple Testing in Genetic Analysis
Author Information
Author(s): Alison P Klein, Ilija Kovac, Alexa JM Sorant, Agnes Baffoe-Bonnie, Betty Q Doan, Grace Ibay, Erica Lockwood, Diptasri Mandal, Lekshmi Santhosh, Karen Weissbecker, Jessica Woo, April Zambelli-Weiner, Jie Zhang, Daniel Q Naiman, James Malley, Joan E Bailey-Wilson
Primary Institution: Inherited Disease Research Branch, NHGRI, NIH, Baltimore, Maryland, USA
Hypothesis
Does the importance sampling method improve the correction for multiple testing in affected sib-pair linkage analysis compared to traditional methods?
Conclusion
The importance sampling method is more efficient than traditional methods for correcting p-values in genetic analysis.
Supporting Evidence
- The importance sampling method showed improved efficiency in estimating p-values compared to traditional methods.
- Type I error rates were conservative across all methods, particularly the Bonferroni method.
- The study demonstrated that the importance sampling method can maintain power while controlling for type I error.
Takeaway
When scientists test many genes at once, they need to make sure they don't get false positives. This study shows a new method that helps them do that better.
Methodology
The study used simulated data to compare the importance sampling method with Bonferroni correction, Feingold method, and naive Monte Carlo simulation in affected sib-pair linkage analysis.
Potential Biases
The limited number of replicates may introduce bias in the conclusions drawn about the methods.
Limitations
The study's power to detect trait loci was generally low, potentially due to the binary trait definitions used.
Statistical Information
P-Value
0.03
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website