Related Experiment Videos
Genetic association tests for family data with missing parental genotypes: a comparison
Alice S Whittemore1, Jerry Halpern
1Department of Health Research and Policy, Stanford University School of Medicine, Stanford, California 94305, USA. alicesw@stanford.de
Genetic Epidemiology
|June 19, 2003
Summary
Comparing genetic association tests in nuclear families, the Rabinowitz test with model-based variance offers good power and robustness. However, its application is limited by complex family structures and varying phenotypes.
Area of Science:
- Genetics
- Statistical genetics
- Bioinformatics
Background:
- Genetic association studies in nuclear families are crucial for identifying disease-related genes.
- Existing statistical tests like Family-Based Association Test (FBAT), Rabinowitz, and Family Genotype Analysis Program (FGAP) have varying assumptions and performance characteristics.
- Evaluating the robustness and power of these tests under different conditions is essential for accurate genetic discovery.
Purpose of the Study:
- To compare the performance of three genetic association tests: FBAT, Rabinowitz, and FGAP.
- To assess the power and type I error rates of these tests, particularly when assumptions are violated.
- To determine the most suitable test for genetic association studies in nuclear families with diverse structures and phenotypes.
Main Methods:
- The study analyzes three genetic association tests: FBAT, Rabinowitz, and FGAP, derived from efficient scores under null expectation constraints.
- Simulations were conducted to evaluate test performance under violated assumptions and with limited family data.
- Power and type I error rates were examined, focusing on scenarios with affected offspring, single-typed parents, and rare, recessive genes.
Main Results:
- FBAT showed less power, especially in families with all affected offspring.
- Rabinowitz and FGAP tests performed similarly, with FGAP extracting more information from families with one typed parent.
- Rabinowitz test with model-based variance demonstrated comparable power to FGAP, with improved type I error rates and robustness, despite computational limitations.
Conclusions:
- The Rabinowitz test with model-based variance is a powerful and robust option for genetic association studies in nuclear families.
- Its practical utility is constrained by the absence of a straightforward algorithm for complex family structures and phenotypes.
- Further development of user-friendly algorithms is needed to enhance the applicability of the Rabinowitz test.