The Study of Coronary Artery Disease Based on Gene and MicroRNA Expression Profiles

• Main Authors: En-Ju Lin, 林恩如
• Other Authors: Austin H. Chen
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/27954695504419509617

碩士 === 慈濟大學 === 醫學資訊學系碩士班 === 101 === Introduction: According to the statistics record from World Health Organization, one third global population dies from heart disease; coronary heart disease (CAD) is one of the main causes. In clinical assessment of heart disease, biomarkers become key criteria. MicroRNAs, ~22 nucleotide non-coding RNAs, regulate negatively to their target genes via degradation or translational inhibition. Recent studies reveal that microRNAs play as important regulators of development and stress responsiveness of heart. Discovering microRNA biomarkers of heart disease is an increasingly aware issue. Methods: In this study, statistic methods, functional similarity scores, and target prediction systems are used to discover the potential microRNA/gene CAD biomarkers and build up the potential CAD related microRNAs-targets visualization network. Moreover, gene algorithm – support vector machine (GASVM) is implemented for the classification of CAD. Results: In this study, 3 out of 5 microRNAs selected via miRTarBase network are verified with the previous biological studies of CAD. In addition to genes involved in cell growth and immunity, genes with function of apoptosis as well as cellular metabolism regulator may also be potential CAD related biomarkers. The selected genes: PNPLA2, FKBP8, SIRT5, DISC1, PBX2, SERPINB8, ADM2, H6PD, ITPK1, MGLL, SLC2A1, JUND, NFAT2IP, and RUNX1, existing the potential connections with CAD need further verification. The visualization network provides the potential microRNA-target pairs, hsa-mir-663-JUND as well as hsa-mir-122-NFATC2IP, for further study. The accuracy of CAD classification with GASVM is up to 89.75% certainly improving the classification performance of the traditional non-invasive diagnosis methods.