Development And Application Of Rice Oligo Array And Phylogenomic Databases

Elucidating the function of all predicted genes in rice is the ultimate goal, while assignment of gene function has progressed slowly, despite the importance of rice and its emergence as a model species for genetic studies. This dissertation concentrates on development of two bioinformatic databases to facilitate gene functional discovery in rice:Rice Oligo Array Database (ROAD) designed to provide a comprehensive gene expression and co-expression profile for all rice genes, and Rice Phylogenomic Database providing a logical and functional format to host and analyze diverse sets of functional genomic data in a phylogenetic context for rice kinases and glycosyltransferases. Based on them,33 rice-diverged GT genes that are highly expressed in above-ground, vegetative tissues, were identified as targets for further functional examination toward understanding and manipulating grass cell wall qualities.This dissertation consists of four chapters.Chapter 1 introduces the current progress of research on rice and rice genome, rice microarray, phylogenomics, cell walls and glycosyltransferases.Chapter 2 represents Rice Oligo Array Database (ROAD) to explore gene expression across 1,867 publicly available rice microarray hybridizations. A user-friendly web interface with several kinds of visualization representation allows users easy extraction of gene expression profiles by gene and microarray element IDs. Meta-analysis is a tool to analyze the expression of genes in anatomy and developmental stage contexts, thereby providing useful information about where and when genes of interest are expressed. Co-expression analysis tool provides information on co-regulation between genes under general, abiotic and biotic stress conditions. Functional analysis tools, including GO and KO enrichment analysis, facilitate identification of meaningful biological patterns in a list of query genes. We applied meta-analysis to evaluate endosperm preferred genes identified from recent study and generated probable functional gene network to explain endosperm development in crop. As a public reference of rice expression database, ROAD will facilitate functional discovery of key genes required for rice production and design of further analysis in rice.Chapter 3 reports the development of two phylogenomic databases for 1,934 rice kinases (Rice Kinase Database, RKD) and 769 glycosyltransferases (Rice GT Database, RGTD). Determination of gene function is particularly problematic when studying large gene families because redundancy limits the ability to experimentally assess the contributions of individual genes. Through RKD and RGTD, several classes of functional genomic data, including mutant lines and gene expression data, can be displayed for each rice kinase and GT in the context of a phylogenetic tree, allowing for comparative analysis both within and between families. This kind of representation, named phylogenomics, at least in part helps us to resolve the gene functional redundancy limitation. Currently, more than 1,000 microarray hybridizations are integrated into these phylogenomic databases and more than 250 protein-protein interactions are integrated into RKD. To show how to use these databases, we describe two applications of RKD for elucidating functions of individual members of rice kinase family:(i) identification of mitogen-associated protein kinase (MAPK-MAPKK-MAPKKK) signaling cascades that are coexpressed in response to a broad range of stresses, and (ii) prediction of five functional interactions of a light-inducible kinase by integrating gene expression patterns with the protein-protein interaction map.Chapter 4 shows how ROAD and RGTD were combined to identify 33 rice-diverged GT genes that are highly expressed in above-ground, vegetative tissues. GTs are critical to the biosynthesis of plant cell walls, among other diverse functions. Based on the CAZy database and sequence similarity searches, we have identified 609 potential GT genes (loci) corresponding to 769 transcripts (gene models) in RGTD. Comprehensive digital expression analysis of public gene expression data from ROAD revealed that most (-80%) rice GTs are expressed. Based on analysis with Inparanoid, we identified 282 "rice-diverged" GTs that lack orthologs in sequenced dicots (Arabidopsis thaliana, Populus tricocarpa, Medicago truncatula and Ricinus communis). Combining these analyses, we identified 33 rice-diverged GT genes (45 gene models) that are highly expressed in above-ground, vegetative tissues. From the literature and this analysis,21 of these loci are excellent targets for functional examination toward understanding and manipulating grass cell wall qualities. Study of the remainder may reveal aspects of hormone and protein metabolism that are critical for rice biology.