Related Experiment Videos
Logistic regression protects against population structure in genetic association studies
Efrosini Setakis1, Heide Stirnadel, David J Balding
1Department of Epidemiology and Public Health, Imperial College, St. Mary's Campus, London W2 1PG, United Kingdom. e.setakis@imperial.ac.uk
Genome Research
|December 16, 2005
Summary
Logistic regression methods effectively protect genetic association studies against false positives from cryptic population substructure, offering a computationally efficient alternative to complex models.
Area of Science:
- Population genetics
- Statistical genetics
- Bioinformatics
Background:
- Cryptic population substructure can lead to false positives in genetic association studies.
- Existing methods like structured association and genomic control have limitations.
Purpose of the Study:
- To compare methods for controlling false positives due to cryptic substructure.
- To evaluate the performance of logistic regression-based approaches.
Main Methods:
- Extensive simulation study.
- Comparison of null marker methods, structured association, genomic control, and logistic regression.
Main Results:
- Structured association methods are effective but computationally intensive.
- Genomic control can reduce statistical power.
- Logistic regression procedures offer good protection without explicit population modeling.
Conclusions:
- Flexible, fast, and easily implemented logistic regression methods provide robust protection against cryptic substructure.
- These methods are a practical alternative for population-based genetic association studies.