Statistical modelling and analysis of Microarray Gene Expression Data

Abstract

DNA microarray experiments raise numerous statistical questions in different fields as diverse as image analysis, experimental design, hypothesis testing, cluster analysis and distribution theory etc. Noise creeps into microarray experiments at each stage from the preparation of tissue samples to the extraction of data. In order to measure gene expression changes accurately, it is important to take into account the random and systematic variations that occur in every microarray experiment. The greatest challenge to array technology lies in the analysis of gene expression data to identify which genes are differentially expressed across tissue samples or experimental conditions. The ability to measure gene expression enmasee has resulted in data with number of variables p far exceeding the number of samples N. Standard statistical methodologies do not work well or even at all when N lt p. Modifications of existing methodologies or development of new methodologies is needed for the analysis of microarray data. Usually in microarray data most genes are expressed at very low levels and only few genes are expressed at high intensity. The main objectives of the research work undertaken were to develop tools specific to microarray data analysis in identification of differentially expressed genes and to have a comparative study with the existing methods, to study the use of statistical classification and dimension reduction techniques in identifying corregulated genes/samples and to study the distribution of gene expression intensities across genes The thesis is organized in six Chapters. Chapter 1 discusses the biological background, design of microarray chip technology and statistical issues in analysis of microarray data.