| Summary: | 博士 === 國立中興大學 === 醫學生物科技博士學位學程 === 106 === The transcriptome contains complete information of all RNAs that are transcribed by the genome at specific developmental stages and under specific physiological or pathological conditions in a particular tissue or cell type. Transcriptomics is a large-scale study of RNA molecules using high-throughput techniques. The process of examining the composition and abundance of transcriptomes in cells by high-throughput techniques is transcriptome profiling. It can be used to identify the expression patterns to classify cellular states or to identify genes with specific expression patterns. DNA microarray analysis and RNA Sequencing (RNA-Seq), a tool of next-generation sequencing (NGS) technologies, have been widely utilized for transcriptome profiling. In this dissertation, we collected, analyzed, and integrated large-scale datasets of transcriptome profiling to study cancer precision medicine, non-invasion diagnosis of disease, plant stress, and crop yield. The architecture of this dissertation is divided into three major components.
The first component is using transcriptome profiling to study cancer characteristics and precision medicine. After collecting and analyzing large-scale cancer RNA-Seq data, we identified a large number of differentially expressed transcripts between different cancer stages or between normal tissues and cancers, which resulted in a comprehensive cancer RNA-Seq database. Additionally, we integrated large-scale cancer expression profiles and drug treatment profiles to develop a novel precision medicine method. The second component is using transcriptome profiling to study the non-invasion diagnosis of disease. Circulating microRNAs (miRNAs) in the blood have been found to be a potential biomarker for non-invasive diagnostics. Although there have been several studies attempting to generate circulating miRNA database, they have not yet integrate the large-scale circulating miRNAs profiles and predict the potential biomarkers using machine learning methods. In this study, we collected and analyzed large-scale DNA microarray analysis of miRNAs and small RNA-Seq data, combined with biological pathways and feature selection methods to construct a circulating miRNA-expression database. The third component is utilizing transcriptome profiling to study the application of transcriptomics in plant stress and crop yield. We collected, processed, analyzed and visualized large-scale publically available plant stress RNA-seq data to construct a plant stress RNA-Seq database. In addition, we also used small RNA-Seq to investigate miRNAs of 129 Taiwanese rice cultivars and examine eight yield-related traits of these rice cultivars. In combination with feature selection and machine learning methods, we have identified certain miRNAs that are closely related to the panicle number traits in rice cultivars.
This study is a framework to provide integrated and comprehensive knowledge for bioinformatics and transcriptome profiling. The results of this study in cancer research provide a resource for research on cancer mechanisms and treatments by biological and medical researchers. The Study on circulating miRNAs will facilitate potential non-invasive biomarkers discovery for routine clinical examinations. The research findings on plant stress and yield may help to solve the food crisis caused by global climate change. Finally, we constructed three large databases, which may contribute to academic interconnection, knowledge base establishment, and mutual validation.
|
|---|
