Database Construction And Bioinformatics Analysis For MicroRNA Biomarker

Micro RNAs(mi RNAs) are small noncoding RNA molecules naturally occurring in eukaryotes, which have become a research hotspot in the biomedical area because of their function as gene expression modulators during development and disease. Many studies have shown that mi RNAs as candidate biomarkers are used for disease or cancer diagnosis, prognosis and treatment response.The contents of this dissertation are divided into two parts. In part one, we developed a high-quality online database to store experimentally supported mi RNA-cancer relationship, thereby to help cancer researchers to get more resource on micro RNA biomarkers. mi R-abc, the manually curated database, aims at providing comprehensive information about micro RNAs biomarkers in human cancers. At present, mi R-abc contains 568 correlations between 228 mi RNAs and 25 types of cancers, which were extracted and curated from 430 papers by processing 2174 publications in Pub Med. mi R-abc is freely available at http://sysbio.suda.edu.cn/mi R-abc/. In part two, we analyzed the human mi RNA-m RNA regulatory network to identify novel micro RNA biomarkers associated with the early diagnosis of sepsis based on the bioinformatics model. By analyzing the mi RNA expression profiles and the mi RNA-m RNA regulatory network, we obtained candidate mi RNAs associated with sepsis. Pathways, ontology enrichment, and protein-protein interaction network(PIN) analyses, as well as ROC curve calculation were exploited to testify the reliability of the prediction. We finally identified 8 novel mi RNAs which have the potential to be the biomarkers for diagnosis of sepsis.