Comparison of statistics for candidate-gene association studies using cases and parents
- Mayo Clinic/Foundation, Rochester, MN (United States)
Studies of association between candidate genes and disease can be designed to use cases with disease, and in place of nonrelated controls, their parents. The advantage of this design is the elimination of spurious differences due to ethnic differences between cases and nonrelated controls. However, several statistical methods of analysis have been proposed in the literature, and the choice of analysis is not always clear. The authors review some of the statistical methods currently developed and present two new statistical methods aimed at specific genetic hypotheses of dominance and recessivity of the candidate gene. These new methods can be more powerful than other current methods, as demonstrated by simulations. The basis of these new statistical methods is a likelihood approach. The advantage of the likelihood framework is that regression models can be developed to assess genotype-environment interactions, as well as the relative contribution that alleles at the candidate-gene locus make to the relative risk (RR) of disease. This latter development allows testing of (1) whether interactions between alleles exist, on the scale of log RR, and (2) whether alleles originating from the mother or father of a case impart different risks, i.e., genomic imprinting. 13 refs., 2 figs., 2 tabs.
- OSTI ID:
- 6872170
- Journal Information:
- American Journal of Human Genetics; (United States), Vol. 55:2; ISSN 0002-9297
- Country of Publication:
- United States
- Language:
- English
Similar Records
The relative efficiency of the Hardy-Weinberg equilibrium-likelihood and the conditional on parental genotype-likelihood methods for candidate-gene association studies
Uniparental disomy of chromosome 16 in offsprings of Familial Mediterranean Fever (FMF) patients treated with colchicine