Expression Database Construction And Co-expression Analysis Of Plant Gene

MicroRNAs (miRNAs) are20-24nucleotide long small non-coding RNAs that play keyregulatory roles during plant development and stress responses. Up to date, a large numberof miRNAs have been identified in a great number of plants as deposited in miRBase andPMRD, with their targets either predicted or validated. Plant genomes contain~100-200miRNAs with an average of2-3targets each that are often transcription factors formingunique regulatory networks.In the past two decades, numerous experiments have been performed using the highdensity gene chips or microarrays. A huge amount of data have been deposited in databasessuch as ArrayExpress and Gene Expression Omnibus (GEO). Much of these data are fromimportant biological experiments such as biotic response, abiotic response, various tissuesand developmental stages and involve many plant species, including important crops such assoybean, rice, and maize.The transcriptome is the complete set of transcripts in a cell, and their quantity, for aspecific developmental stage or physiological condition. Understanding the transcriptome isessential for interpreting the functional elements of the genome and revealing the molecularconstituents of cells and tissues, and also for understanding development and disease. Thekey aims of transcriptomics are: to catalogue all species of transcript, including mRNAs,non-coding RNAs and small RNAs; to determine the transcriptional structure of genes, interms of their start sites,5′and3′ends, splicing patterns and other post-transcriptionalmodifications; and to quantify the changing expression levels of each transcript duringdevelopment and under different conditions.Gene co-expression is the similarity of gene expression patterns under variousexperimental conditions. Prompted by the “guilty by association” principle, i.e., a set ofgenes involved in one biological process are co-regulated and thus co-expressed under thecontrol of a shared regulatory system, gene co-expression analysis has became a powerfultool for gene function prediction during the past5years in various plant species, such asArabidopsis and tomato, and some non-plant species such as yeast and human.There are three main parts of our work: 1. Despite the essential roles of miRNA-target regulation networks in plant stressresponses and development, a well curated, centralized database is not available that one canuse to mine their biological functions by browsing expression patterns with associatedanalytical and visualization tools. Here, we present Plant MiRNA Target ExpressionDatabase (PMTED), a database that comprises a number of querying, analytical, andvisualization functionalities for miRNAs, their targets, and their interaction networks.PMTED is for retrieving expression profiles of miRNA targets represented in the plethora ofexisting microarray data that were manually screened and curated. Besides validatedmiRNAs from available plant miRNA database such as Plant MicroRNA Database (PMRD),expression patterns of targets of novel miRNAs that are defined by the users can also beretrieved via a robust target prediction portal. In additional to a Basic Information extractionfunction for miRNA target sequences, gene ontology, and differential expression profiles,the Meta-Network provides searching and browsing functions for meta-data that weremanually curated and displayed through a Cytoscape Web-based graphical interface. Themeta-networks present a global view about the relationship among species, bioprocesses,conditions, and miRNAs underlined by differentially expressed targets in variousexperiments where novel hypotheses can be established for further experimental validation.PMTED is designed to study the contextual significance of miRNA target genes in theavailable microarray data in the public domain and to assist function network investigation.It should be useful in facilitating the application of these existing data for novel plantresearch projects.2. Wheat is the most important crop in world food production. Triticum aestivum and othercereal genomes are substantially larger than those of A. thaliana and O. sativa. The genomeof hexaploid wheat is about16,000Mb which is38times larger than that of themonocotyledon model O. sativa. It is an allohexaploid composed of three homoeologoussubgenomes, AA, BB, and DD, thus generally each gene is represented by threehomoeologous copies. The large hexaploid nature of the genome is a drawback indetermining the complete sequence which is not yet available.The hexaploid wheat is one of the most important diets for the world. The genomicresearch on wheat, however, is relatively lagged behind because of its gigantic genome size.The study of gene expression data in wheat however is accumulating. Up to date there is nodatabase that is dedicated to wheat expression data. At the time of high throughputsequencing or next generation sequencing (NGS)-based gene expression analysis, it is urgent to have a database that provide functions for wheat expression data collection, integration,and analysis.WhED is a database dedicated to wheat expression data collection, integration, andanalysis. Data types consist of traditional microarray data and newly generated mRNA tagand RNA-Seq data from the next generation sequencing. Functions for array analysis includeprobe search, differential gene screening, GO enrichment analysis, and pathway building.Similar functions were applied to miRNA, siRNA, and RNA-Seq expression data. The twotypes of data were connected through the miRNA targets and the transcripts that generatedthe sRNAs and thus achieve the capability for cross search and comparison between two datatypes. Additional functions include the generation of scatter plot, heat map, and expressiontime course.All data were integrated in GBrowse for integrated visualization.3. To date, a large number of miRNAs have been identified in the model plant Arabidopsisthaliana; however, our understanding of the biological processes regulated by miRNAs isstill limited. Here, we performed a comprehensive analysis of Arabidopsis miRNAs byfocusing on the expression relationships among their targets. We developed a strategy basedon co-expression meta-analysis of miRNA targets (CoMT) to integrate miRNA targetexpression data from hundreds of microarray experiment and analyzed clusters of miRNAtargets leading to the association of miRNAs with biological function(s).