基于突变基因网络的致癌驱动通路检测算法

发布时间：2019-02-28 19:32

【摘要】：通过检测突变驱动通路研究癌症的发病机理是当前癌症基础性研究的关键问题之一.该研究以人类基因组工程提供的体细胞突变数据为研究对象,结合基因组图谱中广泛存在的互斥性原理,提出一种新型的基于基因互斥网络的致癌突变驱动通路检测算法(Megnet).该算法首先利用大量癌症病人的体细胞突变数据,结合基因间互斥性原理构建突变基因网络,然后检测该网络中具有高覆盖的最大完全子图.为验证算法的效率和鲁棒性,我们将该算法应用于模拟数据中,结果显示所有模拟过程均在15秒内完成驱动通路检测,Megnet算法比Dendrix和Multi-Dendrix算法运行时间更短且结果准确率更高.同时为验证算法的有效性,我们将该算法应用于肺癌数据和神经胶质瘤体细胞突变数据中,结果显示Megnet算法不仅比Dendrix和Multi-Dendrix算法检测的基因集合具有更高的生物相关性和统计显著性,而且还检测出一些可供生物验证的新候选基因集合,并且这些检测的基因集合与已知的P53、RB、RAS和PI3K等信号通路及细胞循环和细胞凋亡通路具有较高的重叠.Megnet算法不需要指定通路中的基因个数和任何先验知识,为癌症发病机理研究提供新视野.该算法通过构建突变基因网络,简化了基因间相互关联关系,降低了算法复杂度,提高了致癌突变驱动通路检测的效率和准确性,对于癌症发病机理研究具有较强的理论意义和实践价值.
[Abstract]:To study the pathogenesis of cancer by detecting mutation-driven pathway is one of the key issues in the basic research of cancer. In this study, the somatic mutation data provided by the human genome engineering were taken as the research object, and combined with the mutuality principle widely existed in the genome map, A novel detection algorithm for carcinogenic mutation-driven pathway based on gene mutex network (Megnet).) is proposed in this paper. Based on the somatic mutation data of a large number of cancer patients, a mutagenic gene network was constructed based on the principle of mutual repulsion between genes, and then the maximum complete subgraph with high coverage was detected. In order to verify the efficiency and robustness of the algorithm, we apply the algorithm to the simulation data. The results show that all simulation processes complete the drive path detection within 15 seconds. Megnet algorithm has shorter running time and higher accuracy than Dendrix and Multi-Dendrix algorithm. In order to verify the effectiveness of the algorithm, we apply the algorithm to lung cancer data and glioma cell mutation data. The results showed that Megnet not only had higher biological correlation and statistical significance than the set of genes detected by Dendrix and Multi-Dendrix, but also detected some new candidate gene sets that could be biologically verified. And these detected gene sets overlap with known signaling pathways such as P53, RBB, Ras and PI3K, as well as cell cycle and apoptosis pathways. Megnet algorithm does not need to specify the number of genes in the pathway and any prior knowledge. It provides a new field of view for the study of the pathogenesis of cancer. By constructing a network of mutated genes, the algorithm simplifies the relationship between genes, reduces the complexity of the algorithm, and improves the efficiency and accuracy of detection of carcinogenic mutation-driven pathways. It is of great theoretical and practical value for the study of the pathogenesis of cancer.
【作者单位】：西北农林科技大学信息工程学院
【基金】：中央高校基本科研业务费(2452017342) 博士科研启动费(2452017019) 陕西省自然科学基金面上项目(2017JM6063) 国家自然科学基金重点项目(61532014,61432010) 陕西省杨凌区科技计划项目(2017GY-03)资助~~
【分类号】：Q811.4;R730.2

【相似文献】