Data analysis for microarray experiment and DNA barcode of life

DNA microarray experiment, a well-established experimental technique, aims understanding the function of genes in some biological functions and cellular processes. One of the most common experiments in functional genomic research is to compare two groups of microarray data to determine which genes are differentially expressed. In this dissertation, we propose (1) a methodology to estimate the proportion of differentially expressed genes in microarray experiments, (2) parametric and non-parametric methods to estimate error distribution of microarray data, and (3) an optimal scoring method and LDA on HLdata on the DNA barcoding data to cluster the species using COI sequence. We study the performance of our methods using simulation studies where we compare it to other standard methods and apply it on real data sets to show the advantage of our method.Ph.D.Includes bibliographical references (p. 103-110)