Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


ucin1187009967.pdf (1.35 MB)

ETD Abstract Container

Abstract Header

Some Statistical Aspects of Association Studies in Genetics and Tests of the Hardy-Weinberg Equilibrium

He, Ran
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1187009967

Abstract Details

2007, PhD, University of Cincinnati, Medicine : Biostatistics (Environmental Health).
The applicability of a statistical method hinges how far the assumptions are met for its validity. Some statistical tests are robust when assumptions are relaxed, while others are not. In first part of the dissertation, we focus on exploring assumption violations in some statistical methods for association studies in Genetics and use simulations to test the robustness of these methods. In genetic studies, one of the major objectives is to apply statistical models to identify genes contributing to variations in specific quantitative traits. The method of analysis of variance (ANOVA) is frequently used. However, we established that this method raises a paradox by violating the assumptions of its validity. An alternative method, namely Bartlett’s test, is available to overcome the paradox. We compare the performances of the ANOVA test and Bartlett’s test to answer the underlying question. Our study indicates that the ANOVA test works despite the failure of the validity of its assumption. In the second part of the dissertation, we focus on tests of the Hardy-Weinberg Equilibrium (HWE). The frequently used method is the goodness-of-fit Chi-squared statistic, which does not discriminate homozygote excess from heterozygote deficiency because of its two-sided nature. We propose alternative methods and use simulations to assess their power. The proposed methods are amenable to sample size calculations. We compare our sample size calculations with those available in the literature and find that ours are smaller. For more than two alleles, testing the HWE is computationally complex. We propose a new method of testing the HWE for multi-allelic cases by reducing the dimensionality of the problem. Mathematical, statistical, and computational aspects of the new method are set out in great detail.
Dr. Marepalli Rao (Advisor)
155 p.
Biology, Biostatistics
association studies in Genetics; paradox; assumption violation; tests of HWE; dimension reduction; simulations

Recommended Citations

Citations

  • He, R. (2007). Some Statistical Aspects of Association Studies in Genetics and Tests of the Hardy-Weinberg Equilibrium [Doctoral dissertation, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1187009967

    APA Style (7th edition)

  • He, Ran. Some Statistical Aspects of Association Studies in Genetics and Tests of the Hardy-Weinberg Equilibrium. 2007. University of Cincinnati, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1187009967.

    MLA Style (8th edition)

  • He, Ran. "Some Statistical Aspects of Association Studies in Genetics and Tests of the Hardy-Weinberg Equilibrium." Doctoral dissertation, University of Cincinnati, 2007. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1187009967

    Chicago Manual of Style (17th edition)

ucin1187009967
611
© 2007, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.