| Summary: | 碩士 === 輔仁大學 === 應用統計學研究所 === 97 === Family-based association tests are widely used in gene mapping. Pedigree disequilibrium test (PDT) is one of these tests for analysis of pedigree data. PDT derived by combining transmission disequilibrium and allele difference. In this thesis, we tried to combine the information of odds ratio, similarity difference, transmission disequilibrium and allele difference. We use some existing combination procedures, such as Tippett, Fisher and inverse normal method to combine the four statistics information. Overall, we have considered 48 statistics for investigating which of them having higher power performances. The simulation result shows that the statistic constructed with similarity difference has better power performance than other statistics.
|
|---|
