...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Integrative Bioinformatics >Discovery of miR-mRNA interactions via simultaneous Bayesian inference of gene networks and clusters using sequence-based predictions and expression data
【24h】

Discovery of miR-mRNA interactions via simultaneous Bayesian inference of gene networks and clusters using sequence-based predictions and expression data

机译：使用基于序列的预测和表达数据通过基因网络和簇的同时贝叶斯推理发现miR-mRNA相互作用

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

MicroRNAs (miRs) are known to interfere with mRNA expression, and much work has been put into predicting and inferring miR-mRNA interactions. Both sequence-based interaction predictions as well as interaction inference based on expression data have been proven somewhat successful; furthermore, models that combine the two methods have had even more success. In this paper, I further refine and enrich the methods of miRmRNA interaction discovery by integrating a Bayesian clustering algorithm into a model of prediction-enhanced miR-mRNA target inference, creating an algorithm called PEACOAT, which is written in the R language. I show that PEACOAT improves the inference of miR-mRNA target interactions using both simulated data and a data set of microarrays from samples of multiple myeloma patients. In simulated networks of 25 miRs and mRNAs, our methods using clustering can improve inference in roughly two-thirds of cases, and in the multiple myeloma data set, KEGG pathway enrichment was found to be more significant with clustering than without. Our findings are consistent with previous work in clustering of non-miR genetic networks and indicate that there could be a significant advantage to clustering of miR and mRNA expression data as a part of interaction inference.
机译：已知MicroRNA（miR）会干扰mRNA表达，并且已经在预测和推断miR-mRNA相互作用方面进行了大量工作。基于序列的交互预测以及基于表达数据的交互推断都已被证明是成功的。此外，结合两种方法的模型取得了更大的成功。在本文中，我通过将贝叶斯聚类算法集成到预测增强型miR-mRNA目标推论模型中，创建了一种名为PEACOAT的算法，该算法以R语言编写，从而进一步完善和丰富了miRmRNA相互作用的发现方法。我表明，PEACOAT可以使用模拟数据和多发性骨髓瘤患者样品微阵列数据集来改善miR-mRNA靶标相互作用的推断。在包含25个miRs和mRNA的模拟网络中，我们使用聚类的方法可以在大约三分之二的情况下改善推断，并且在多发性骨髓瘤数据集中，发现聚簇比不使用KEGG途径富集更为重要。我们的发现与以前在非miR遗传网络聚类中的工作一致，并表明将miR和mRNA表达数据聚类作为交互推理的一部分可能具有显着优势。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号