论文部分内容阅读
目的:基于基因芯片数据挖掘的方法，对小儿肝母细胞瘤预后差异基因进行筛选后，对差异基因的功能及通路进行分析，同时利用共表达网络筛选出决定小儿肝母细胞瘤预后的核心基因并对其预测能力进行评估。方法:本研究中所使用的小儿肝母细胞瘤基因芯片表达谱来自欧洲生物信息学研究所(http：//www.ebi.ac.uk/embl/)；数据收集的截止日期为2018年12月31日。先通过SAM法筛选得到差异表达基因(基因表达水平上升至原来的2倍或下降至原来的1/2)，再基于降维原理通过共表达网络模型筛选得到核心基因，运用MCODE算法计算基因的调控评估能力评分，以评价其在整个网络模型中的调控能力。结果:针对213个差异表达基因的细胞信号通路富集结果显示，富集度最高的信号通路为代谢途径通路，富集度为2 122.529，信号通路富集结果误判率均<0.001，经SAM算法检验均具有统计学意义(n P<0.001)。以213个在不同预后组发生差异表达的基因作为共表达网络的构建基础，本次构建得到的共表达网络共纳入12个发生差异表达的基因。以预后不良组为实验组，以预后较好组为对照组，采用MCODE算法计算基因调控能力评分的结果显示，决定小儿肝母细胞瘤预后调控能力评分最高基因为ADH1A基因，得分为19分。此外，HAO1、ADH1B、ALDOB以及DPYS基因的调控能力评分均高于或接近于5分，因此可认为它们是本次构建得到的共表达网络模块中的核心基因。n 结论:从共表达网络模型结果来看，ADH1A基因与小儿肝母细胞瘤的发生发展存在较为密切的关联，其分子生物学证据对临床开展肿瘤靶向干预疗法的指导意义有待进一步挖掘。“,”Objective:Based on the microarray data mining method, the function and pathway of differential genes were analyzed after the differential genes were screened. At the same time, the core genes that determine the prognosis of pediatric hepatoblastoma were screened by coexpression network, and their predictive ability was evaluated.Methods:The microarray expression profile of pediatric hepatoblastoma used in this study was from the European Institute of bioinformatics (http: //www.ebi.ac.uk/embl/). The deadline for data collection was December 31, 2018. Firstly, the differentially expressed genes (gene expression level increased to 2 times or decreased to 1/2 of the original) were screened by SAM method, then the core genes were screened by coexpression network model based on dimension reduction principle, and the gene regulation evaluation score was calculated by MCODE algorithm to evaluate its regulation ability in the whole network model.Results:According to the enrichment results of 213 differentially expressed genes, the highest enrichment degree of signal pathway was metabolic pathways (2 122.529). The misjudgment rate of signal pathway enrichment results was less than 0.001, and the misjudgment rate was statistically significant by SAM method (n P<0.001). A total of 213 differentially expressed genes in different prognosis groups were used as the basis for the construction of the coexpression network. A total of 12 differentially expressed genes were included in the coexpression network. Using the poor prognosis group as the experimental group, and the better prognosis group as the control group, the MCODE algorithm was used to calculate the gene regulatory ability score. The results showed that the highest gene for determining the prognosis control ability of children hepatblastoma was ADH1A gene with a score of 19. In addition, the regulatory ability scores of HAO1, ADH1B, ALDOB and DPYS genes were higher than or close to 5, so they could be considered as the core genes in the coexpression network module.n Conclusions:According to the results of coexpression network model, ADH1A gene is closely related to the occurrence and development of hepatoblastoma in children, and its molecular biological evidence needs to be further explored to guide the clinical development of tumor targeted intervention therapy.