| Summary: | 博士 === 國立臺灣大學 === 流行病學研究所 === 93 === Family-based association study is a useful approach for detecting linkage and linkage disequilibrium between a disease gene and a marker in genetic analysis. Using case-parent data the transmission/disequilibrium test (TDT) method tests the equality of the transmission and non-transmission probabilities of a particular maker allele from a heterozygous parent. If the null hypothesis is rejected, it indicates that there is linkage and linkage disequilibrium between the studied disease and the marker loci. Though TDT has been shown as an applicable method in linkage study, its test performance is affected by some nuisances such as gene frequencies, mode of inheritance (MOI) of a disease and linkage disequilibrium between loci. For instance, when the disease allele frequency is low and the disease is recessivity inherited , the power performance of TDT is poor. Theoretically, if we know the actual MOI, the MOI can be involved in the analysis. However, in practice we usually have no idea about the disease mode, and that causes a problem that needs to be addressed. Using robust methods in resoling this problem for a binary trait had been studied. In this paper, we are interested in developing robust methods for handling MOI problem for quantitative traits. By means of conditional likelihoods, we constructed score tests for four modes of inheritance. We then developed two robust procedures to cope with the MOI problem in analysis of quantitative traits. Simulation results showed that our methods truly express robustness property when the actual MOI is unknown.
|
|---|
