Drosophila regulatory genomics: cis-regulatory modules, promoters, and microarrays
MetadataShow full item record
For all biological organisms, when, where and how much a gene is expressed is under sophisticated regulation through cis - and trans -acting factors. Using Drosophila as a model system, I investigated the regulation of gene expression by cis -acting factors including cis -regulatory modules (CRMs) and promoters, and the profiling of gene expression by whole-genome microarrays. CRMs, including enhancers, silencers, insulators, refer to the DNA regulatory elements that are responsible for modulating gene expression relative to the basal transcription level. In chapter 2, I incorporate conservation information into genome-wide CRM prediction and successfully identify three truly functional CRMs, while using another approach of selecting CRMs whose flanking genes share high protein sequence similarity does not work well. In chapter 3, I show the widely used Stinger series of vectors alone can drive reporter gene expression in vivo. In chapter 4, I summarize the findings from a systematic investigation of CRM features. Experimentally verified CRMs are found to be more GC-rich, more conserved across evolution, and more likely to be transcribed than random non-coding sequences. However the differences are marginal and the dense clustering of transcription factor binding sites, which has been applied in many CRM prediction studies, is not a common feature among all CRMs. In chapter 5, I focus on the genome wide organization of another class of cis -acting factors--the promoters. My results suggest that rather than simply an assembly station of the basal transcription machinery, the promoters play an active role in coordinating gene expression between nearby genes and between multiple transcripts of the same gene. Microarrays are a powerful tool to capture gene expression profiles within biological samples. In order to identify the best way of analyzing microarray data, I constructed an improved wholly defined spike-in dataset and used it to assess the various methods proposed for microarray data analysis. As shown in Chapter 6, my study pointes out whether changes in gene expression are balanced or not affect how the data should be analyzed. Based on this finding, I propose a new procedure for microarray analysis, wherein balanced versus unbalanced gene expression is diagnosed first, and then the corresponding best route is performed to achieve the highest sensitivity and specificity in the detection of differentially expressed genes.