加权基因共表达网络分析（WGCNA）探索肺腺癌中的功能基因模块

发布时间：2018-05-03 03:20

本文选题：肺腺癌 + 加权基因共表达网络分析　；参考：《北京协和医学院》2017年博士论文

【摘要】：背景与目的肺腺癌(Lung adenocarcinoma,LUAD)是最为多发的肿瘤之一,其发病率及死亡率在我国及全世界均居前列。研究肺腺癌的分子机制对于早期诊治及预后改善具有重要意义。本研究意在使用网络分析的方法分析肺腺癌基因表达谱,以期获得与临床特性及生存时长有显著关系的基因模块,并对模块基因进行进一步挖掘。方法本研究使用加权基因共表达网络分析(Weighted gene co-expression network analysis,WGCNA)的方法,对56例LUAD肿瘤样本的3513个基因的表达谱构建网络(网络A)进行分析,构建基因模块并验证其与肿瘤TNM分期,病理分级以及生存数据的相关性;对56例LUAD肿瘤样本及相应的正常组织样本,构建联合加权基因共表达网络(网络B)并分析其3513个基因的表达谱,构建基因模块并验证其与样本种类的相关性;将网络A及网络B联合,分析其基因模块重合度,并对重合度高的两对基因模块内的基因进行Gene Ontology及KEGG通路分析。结果WGCNA算法在网络A中划分出12个基因模块,其中1个与患者肿瘤M分期呈中等强度显著相关(R20.4,p0.001);经Cox回归生存分析发现,其中4个与患者发生死亡事件的风险显著相关(p0.05)。在网络B中划分出13个基因模块,其中9个均与样本种类呈强相关(R20.8,p0.001)。对高重合度的两对基因模块,其Gene Ontology及KEGG通路分析均显示有若干生物过程、分子功能及细胞组分的词条明显富集,提示同一基因模块对中的蛋白编码基因及长非编码RNA基因在相应方面可能具有相似性。结论本研究表明WGCNA所划分的基因模块具有生物学意义,且联合临床及样本信息亦能验证模块的临床意义,能够作为探索LUAD的分子机制,发现潜在的基因生物标志物的方法学工具。
[Abstract]:Background & objective Lung adenocarcinoma LUADA is one of the most frequent tumors, and its morbidity and mortality are among the highest in China and the world. It is important to study the molecular mechanism of lung adenocarcinoma for early diagnosis, treatment and prognosis improvement. The purpose of this study is to analyze the gene expression profile of lung adenocarcinoma by network analysis, in order to obtain the gene module which has significant relationship with clinical characteristics and survival time, and further excavate the gene of the module. Methods in this study, a weighted gene co-expression network analysis method was used to analyze the expression profiles of 3513 genes in 56 cases of LUAD tumors. The gene modules were constructed and their association with tumor TNM staging was verified. The correlation of pathological grading and survival data, 56 cases of LUAD tumor samples and corresponding normal tissue samples, constructed a combined weighted gene coexpression network (network B) and analyzed the expression profiles of 3513 genes. Construct the gene module and verify its correlation with the sample type; combine network A and network B to analyze the coincidence degree of the gene module, and analyze the Gene Ontology and KEGG pathway of the two pairs of genes in the gene module with high overlap. Results the WGCNA algorithm was divided into 12 gene modules in network A, one of which was significantly correlated with tumor M stage (R20.4p0.001), 4 of which were significantly correlated with the risk of death events by Cox regression analysis (P 0.05). In network B, 13 gene modules were divided, of which 9 were strongly correlated with the sample species R20.8p0.001. For the two pairs of gene modules with high coincidence degree, the Gene Ontology and KEGG pathway analysis showed that there were several biological processes, and the terms of molecular function and cell components were obviously enriched. It is suggested that the protein encoding genes and the long non-coding RNA genes in the same gene module pair may be similar in the corresponding aspects. Conclusion this study indicates that the gene modules divided by WGCNA have biological significance, and the combination of clinical and sample information can also verify the clinical significance of the modules, which can be used as molecular mechanisms to explore the LUAD. Methodological tools for identifying potential genetic biomarkers.
【学位授予单位】：北京协和医学院
【学位级别】：博士
【学位授予年份】：2017
【分类号】：R734.2

【相似文献】