Related Experiment Videos
Testing for genetic heterogeneity in the genome search meta-analysis method
Cathryn M Lewis1, Douglas F Levinson
1Department of Medical and Molecular Genetics, King's College London School of Medicine at Guy's, King's College and St. Thomas' Hospitals, Guy's Hospital, London, UK. Cathryn.lewis@genetics.kcl.ac.uk
Genetic Epidemiology
|April 6, 2006
Summary
The rank-restricted test is recommended for Genome Search Meta-Analysis (GSMA) heterogeneity testing, despite low power to detect heterogeneity. The rank-unrestricted test shows biased results for linkage detection.
Area of Science:
- Genetics
- Statistical genetics
- Bioinformatics
Background:
- Genome Search Meta-Analysis (GSMA) aggregates genome-wide linkage study results using summed rank statistics.
- Existing heterogeneity testing methods (Q, Ha, B) by Zintzaras and Ioannidis (2005) were evaluated.
- Two testing procedures, rank-restricted and rank-unrestricted, were analyzed for their performance.
Discussion:
- The rank-unrestricted test exhibits conservative behavior for high heterogeneity and liberal behavior for low heterogeneity in linked regions.
- The rank-restricted test is suggested for use, acknowledging the need for extensive simulations.
- Simulation studies reveal low power to detect heterogeneity, even with substantial linkage evidence.
Key Insights:
- The Q statistic showed low power (14-29%) for detecting heterogeneity in simulated affected sib pair studies.
- Despite low heterogeneity detection power, the summed rank statistic demonstrated high power (79-98%) for linkage detection.
- Current heterogeneity testing in GSMA offers limited added value beyond the summed rank statistic.
Outlook:
- Further research may refine heterogeneity testing methods within GSMA.
- Improved simulation strategies could enhance the power of heterogeneity detection.
- The findings emphasize the robustness of the summed rank statistic for primary linkage detection.