Open Daily: 10am - 10pm | Alley door pickup 10am - 7pm 3038 Hennepin Ave Minneapolis, MN
612-822-4611

  • Open Daily: 10am - 10pm
    Alley-side Pickup: 10am - 7pm

    3038 Hennepin Ave Minneapolis, MN
    612-822-4611

Open Daily: 10am - 10pm | Alley-side Pickup: 10am - 7pm
3038 Hennepin Ave Minneapolis, MN
612-822-4611
Restricted Parameter Space Models for Testing Gene-Gene Interaction.

Restricted Parameter Space Models for Testing Gene-Gene Interaction.

Song, Minsun

Paperback

Biology

Currently unavailable to order

ISBN10: 1243520841
ISBN13: 9781243520845
Publisher: Proquest Umi Dissertation Pub
Pages: 92
Weight: 0.40
Height: 0.19 Width: 7.44 Depth: 9.69
Language: English
In this thesis, we present statistical methods to address problems in studying interactions between two unlinked markers. There is a growing recognition that interactions (gene-gene and gene-environment) can play an important role in common disease etiology. The development of cost-effective genotyping technologies has made genome-wide association studies the preferred tool for searching for loci affecting disease risk. These studies are characterized by a large number of investigated SNPs, and efficient statistical methods are even more important than in classical association studies that are done with a small number of markers. In this thesis we propose a novel gene-gene interaction test that is more powerful than classical methods. The increase in power is due to the fact that the proposed method incorporates reasonable constraints in the parameter space. The test for both association and interaction is based on a likelihood ratio statistic that has a chi-bar-squared distribution asymptotically. We also discuss the definitions used for no interaction and argue that tests for pure interaction are useful in genome-wide studies, especially when using two stage strategies where the analyses in the second stage are done on pairs of loci for which both SNPs have at least moderate evidence for associations. We also consider the problem of testing interactions for untyped markers (unobserved variants). Reference databases such as that from the HapMap International Project are providing a great deal of information on typed and untyped markers in study sample for various human populations and this can be used for imputing missing data. We consider a two-stage procedure in which probabilities of genotype assignments are estimated via the imputation method and these probabilities are used as pseudo-observations in the analysis. A Wald test and a generalized Wald test are proposed, both using the jackknife for taking into account the uncertainty in predicting the genotype at the untyped markers. Simulation studies have been conducted to compare the performance within the tests for untyped markers and between the tests for typed variants and for untyped variants. We find that the tests we propose for untyped markers are valid and perform reasonably in terms of power. In the last part of the thesis, we consider the problem of measuring relative information for studying gene-gene interactions. After analysis for interaction between untyped SNPs, one needs to design follow-up studies. Our goal is to decide whether to type the markers which show moderate significance or to determine the number of subjects which are needed to be genotyped for the same level of significance for the purpose of validation. Based on asymptotic relative efficiency, we propose a measure which (1) is easy to compute and (2) is connected in some situations to the measure of Nicolae et al. (2008), which conditions on particular data sets. Simulation studies show that our measure is a reliable index of relative information. Application to a Crohn's disease dataset is presented in the last chapter.

Also in

Biology

View All