论文部分内容阅读
目的:寻找新的卵巢癌发病相关microRNA并为研究人员提供优化后的卵巢癌风险microRNA参照列表。方法:通过在生物网络中度量microRNA靶基因与卵巢癌基因间的功能相似性设计并实现优化卵巢癌风险microRNA计算学方法。采用留一法交叉证实检测该方法的准确性。应用该方法对人类1 527个microRNA进行优化排序。结果:留一法交叉证实所得ROC曲线下面积0.92,该方法有着较高的灵敏度和特异度。排序后,一些已知的卵巢癌相关microRNA如let-7、miR-34/200排在了优化结果的前20位。与新一代测序数据结果进行比较,发现排序前20位microRNA中的大部分都在正常和卵巢癌组织中呈差异表达。结论:应用计算学方法可筛选出卵巢癌相关microRNA,并提供优化后的风险microRNA列表。miR-449a等7个未被报道与卵巢癌有关的miRNA有望成为新的卵巢癌相关的风险因子。
OBJECTIVE: To search for new microRNAs associated with the pathogenesis of ovarian cancer and to provide researchers with a refined microRNA reference list of ovarian cancer risk. METHODS: Design and implementation of microRNA computational methods to optimize ovarian cancer risk by measuring functional similarity between microRNA target genes and ovarian cancer genes in biological networks. The left-behind method was used to confirm the accuracy of the method. This method is used to optimize the sequencing of 1 527 microRNAs in humans. Results: The area under the ROC curve obtained by one-way crossover test was 0.92. This method has high sensitivity and specificity. After sorting, some of the known ovarian cancer-associated microRNAs such as let-7 and miR-34/200 were ranked in the top 20 of the optimization results. Compared with the results of the new generation of sequencing data, most of the 20 microRNAs in the rankings were found to be differentially expressed in normal and ovarian cancer tissues. Conclusion: Computational methods can be used to screen ovarian cancer-associated microRNAs and provide a list of optimized microRNAs. Seven non-reported miRNAs associated with ovarian cancer such as miR-449a are expected to be the new risk factors associated with ovarian cancer.