Masters Theses

Abstract

"Microarrays are a new technology to investigate the expression levels of thousands of genes simultaneously. Microarrays present new statistical problems because the data are very high dimensional with very little replication. Identifying differentially regulated genes in experiments comparing two experimental conditions is often a key step in the microarray data analysis process.

In this thesis different approaches and methodological developments are investigated. The null hypothesis that a gene j is not differentially expressed in two experimental conditions by H0: µ1j = µ2j; versus the alternative, H0: µ1j; :f. µ2j; is tested using three types oft-tests, one using pooled variance and degrees of freedom, one using a Welch correction on the degrees of freedom, and another using an ANOV A model for estimation of standard errors. Comparison of the distributions of p-values from the three methods are discussed with respect to a microarray dataset as well as simulated datasets.

The second key area of investigation is a global null hypothesis that there are no genes differentially expressed in a microarray experiment. The performance of an F-test on a gene-treatment interaction term in an ANOV A mixed model is investigated, and a residual bootstrap method is considered as an alternative to using the parametric F distribution. A simulation study is carried out under various assumptions about the distribution of the data. Performance of the ANOV A F-test method and bootstrap technique are evaluated with regard to type I error and power"--Abstract, page iii.

Advisor(s)

Gadbury, Gary L.

Committee Member(s)

Samaranayake, V. A.
Xing, Xuejing

Department(s)

Mathematics and Statistics

Degree Name

M.S. in Applied Mathematics

Publisher

University of Missouri--Rolla

Publication Date

Summer 2005

Pagination

viii, 71 pages

Note about bibliography

Includes bibliographical references (pages 68-70)

Rights

© 2005 Wanrong Yin, All rights reserved.

Document Type

Thesis - Restricted Access

File Type

text

Language

English

Subject Headings

DNA microarrays -- Statistical methodsGene expression -- Research -- MethodologyBootstrap (Statistics)Analysis of variance

Thesis Number

T 8805

Print OCLC #

66899772

Download
Off-campus S&T users

Share My Thesis If you are the author of this work and would like to grant permission to make it openly accessible to all, please click the button above.

Share

 
COinS