Laboratory of Systems Biology and Bioinformatics (LBB), Institute of Biochemistry and Biophysics, University of Tehran, Tehran, Iran.
Centre for Computational Biology, Haworth Building, University of Birmingham, Birmingham, UK.
J Cell Biochem. 2019 Apr;120(4):5459-5471. doi: 10.1002/jcb.27825. Epub 2018 Oct 9.
Understanding the genetic causes of neurodegenerative disease (ND) can be useful for their prevention and treatment. Among the genetic variations responsible for ND, heritable germline variants have been discovered in genome-wide association studies (GWAS), and nonheritable somatic mutations have been discovered in sequencing projects. Distinguishing the important initiating genes in ND and comparing the importance of heritable and nonheritable genetic variants for treating ND are important challenges. In this study, we analysed GWAS results, somatic mutations and drug targets of ND from large databanks by performing directed network-based analysis considering a randomised network hypothesis testing procedure. A disease-associated biological network was created in the context of the functional interactome, and the nonrandom topological characteristics of directed-edge classes were interpreted. Hierarchical network analysis indicated that drug targets tend to lie upstream of somatic mutations and germline variants. Furthermore, using directed path length information and biological explanations, we provide information on the most important genes in these created node classes and their associated drugs. Finally, we identified nine germline variants overlapping with drug targets for ND, seven somatic mutations close to drug targets from the hierarchical network analysis and six crucial genes in controlling other genes from the network analysis. Based on these findings, some drugs have been proposed for treating ND via drug repurposing. Our results provide new insights into the therapeutic actionability of GWAS results and somatic mutations for ND. The interesting properties of each node class and the existing relationships between them can broaden our knowledge of ND.
了解神经退行性疾病（ND）的遗传原因对于其预防和治疗可能是有用的。在导致 ND 的遗传变异中，全基因组关联研究（GWAS）中发现了可遗传的种系变异，而在测序项目中发现了不可遗传的体细胞突变。区分 ND 中的重要起始基因，并比较可遗传和不可遗传遗传变异在治疗 ND 中的重要性是重要的挑战。在这项研究中，我们通过考虑随机网络假设检验程序进行基于有向网络的分析，对来自大型数据库的 ND 的 GWAS 结果、体细胞突变和药物靶点进行了分析。在功能相互作用组的背景下创建了一个疾病相关的生物网络，并解释了有向边类的非随机拓扑特征。层次网络分析表明，药物靶点往往位于体细胞突变和种系变异的上游。此外，我们利用有向路径长度信息和生物学解释，提供了这些创建节点类中最重要基因及其相关药物的信息。最后，我们确定了九个与 ND 药物靶点重叠的种系变异，七个来自层次网络分析的接近药物靶点的体细胞突变，以及网络分析中控制其他基因的六个关键基因。基于这些发现，一些药物已被提议通过药物再利用来治疗 ND。我们的结果为 ND 的 GWAS 结果和体细胞突变的治疗可操作性提供了新的见解。每个节点类的有趣性质和它们之间的现有关系可以拓宽我们对 ND 的认识。
