Abstract:
|
Genomic imprinting and maternal effects are two epigenetic factors increasingly explored for their roles in the etiology of complex diseases. Full likelihood based statistical methods have been proposed to detect these two effects simultaneously. Such methods, however, have to make strong assumptions concerning mating type probabilities to avoid overparameterization. In this talk, we describe a partial Likelihood method (LIME) that is applicable to a mixture of case-parent/control-parent triads and case-mother/control-mother pairs. Data from additional siblings may also be included. By matching case families with control families of the same structure and stratifying according to familial genotypes, one can derive a partial likelihood that is free of the nuisance parameters. This renders unnecessary any unrealistic assumptions and leads to a robust procedure without sacrificing power. Simulation study demonstrates that LIME has correct type I error rate, little bias and reasonable power under a variety of settings. Based on asymptotic properties of LIME, we investigate severalstudy designs regarding expected information, leading to recommendations on an efficient study design.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.