基于基因关系网络的单克隆抗体靶向抗肿瘤药物的药物警戒研究

发布时间：2018-03-20 22:12

本文选题：生物信息学　切入点：基因　出处：《第二军医大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,由于恶性肿瘤发病率和肿瘤全球化趋势不断上升,人类的生命健康安全饱受威胁,而传统的化疗药物在癌症治疗中存在耐药性、疗效低且毒副作用大等各种缺陷。进入二十一世纪以来,随着生物技术的进步和对肿瘤机制的深入探索,靶向抗肿瘤疗法开始应用于临床,成为现代医学中最强大的治疗和诊断工具,在肿瘤学领域也变得越来越重要。单克隆抗体靶向抗肿瘤药物本质上是一类修饰蛋白,以肿瘤细胞特定部位为靶点,专门抑制在肿瘤生长中的信号通路,诱导肿瘤细胞产生免疫应答,从而选择性杀伤肿瘤细胞,相对于传统化疗药物有着高效低毒的优势,但随着临床使用增加,也表现出各种副反应症状,如胃肠道毒性、心血管毒性和皮肤毒性反应等。目前对于单抗类靶向抗肿瘤药物的研究也主要是临床用药的案例报道和相关文献综述分析,很少就单抗药物不良反应的发生机制进行探究。有研究表明,基因和不良反应之间有一定的关联性,然而关于此类研究主要是借助药物流行病学研究或分子生物学研究进行,往往投入大量人力物力财力。当前已有生物信息学课题组开始进行基因相关的药物不良反应研究,但主要以收集所有上市药物并建立相关模型和数据库为主,缺乏对某类药物的具体研究。研究目的:鉴于以上研究现状,本研究采用生物信息学手段对单克隆抗体靶向抗肿瘤药物的基因--不良反应关联性进行挖掘研究,意在探索基因--不良反应关联关系研究的新思路,为分子生物学和药物流行病学的进一步研究提供理论基础,为单克隆抗体靶向抗肿瘤药物的不良反应研究提供初步参考。研究方法:本研究通过文献检索、数据库查询,收集单抗类药物--基因作用信息和药物--不良反应关联关系数据,来构建药物--基因--不良反应关联关系网络,然后联合关联规则挖掘算法和频数法(ROR、PRR、χ2、MHRA和Yule’s Q)进行信号挖掘,根据挖掘结果筛选出高关联信号进行深入分析。数据统计采用EXEL 2010进行计算。研究结果:1、本研究共纳入14个单克隆抗体靶向抗肿瘤药物,收集到药物--基因相互作用信息记录638条,药物--不良反应数据记录1151条,构建一一对应的基因--药物--不良反应网络共60258条。2、关联规则算法作用度设为Lift2时,共检出信号829个,相对Yule’s Q、PRR、ROR、χ2、MRHA检出信号重合率分别为73.95%、57.39%、57.39%、24.65%、3.81%.3、关联规则算法挖掘结果筛选出4个基因--不良反应关联信号较强的药物,分别是阿柏西普,派姆单抗,纳武单抗,西妥昔单抗。4、筛选关联规则算法和频数法(Yule’s Q、PRR、ROR、χ2)挖掘结果中信号强度和重合度较高的基因--不良反应配对结合相关的药物进行分析,分别是阿柏西普各部位出血不良反应研究、派姆单抗皮肤不良反应研究和西妥昔单抗呼吸系统不良反应研究。研究结论:本研究构建了单克隆抗体靶向抗肿瘤药物--基因--不良反应关联关系网络,联合运用关联规则挖掘算法和频数法进行信号挖掘;对比关联规则算法和频数法信号挖掘结果,发现两类方法的重合度较好;筛选出4个基因--不良反应关联信号较强的药物;通过文献检索和数据库查询,对高关联信号进行分析,发现信号挖掘结果对后续研究有一定的参考性;本研究属于基础研究,为药物流行病学研究和单克隆抗体靶向抗肿瘤药物不良反应的进一步研究提供了初步的数据参考和理论基础。
[Abstract]:In recent years, the trend of malignant tumor incidence and tumor globalization rising, threatened human health and safety, while the traditional chemotherapy drugs resistance existed in the treatment of cancer, curative effect and low toxic side effects and other defects. Since twenty-first Century, with the progress of biotechnology and to explore the mechanism of tumor target. Begin to be applied in clinical anti-tumor therapy, treatment and become the most powerful diagnostic tools in modern medicine, has become more and more important in the field of oncology. Monoclonal antibody targeted anticancer drugs are essentially a class of modified proteins to specific parts of tumor cell targeting inhibition in the pathway of tumor growth specifically, the immune response induced by tumor cells, thereby selectively killing tumor cells, compared with the traditional chemotherapy drugs have the advantages of high efficiency and low toxicity, but with the increase of clinical use, Also exhibit various side effects such as gastrointestinal symptoms, toxicity, cardiovascular toxicity and skin toxicity. The monoclonal antibody targeting of anticancer drugs is mainly clinical case report and literature review analysis, little adverse reaction mechanism of monoclonal antibody drugs are explored. Studies have shown that there is association certain between genes and adverse reactions, however, about this kind of research is mainly carried out by the study of epidemiological studies of molecular biology or medicine, often put a lot of manpower and material resources. The current bioinformatics research group began to study the adverse drug reaction related genes, but mainly to collect all the listed drugs and establish the related model and database. The lack of specific research, for a certain class of drugs. Objective: in view of the above research, this study uses bioinformatics means of single Research on mining cloning antibody targeted antitumor drug adverse reaction gene relevance, to explore new ideas for gene -- Study on the relationship between the adverse reactions, and provide a theoretical basis for further research on the molecular biology and drug epidemiology, provide preliminary reference for monoclonal antibody targeting study of adverse reactions of antitumor drugs. Methods: This study through literature search, database query, collection of monoclonal antibody drugs -- gene information and drug adverse reaction -- correlation data, to construct the drug adverse reaction gene association network, then the joint association rule mining algorithm and frequency method (ROR, PRR, MHRA and Yule was 2, s Q) signal according to the mining, mining results screened high correlation signal were analyzed. Data were analyzed by EXEL 2010 were calculated. Results: 1. This study included 14 monoclonal antibody Body targeted anticancer drugs, collected drug -- gene interaction information records 638, drug adverse reaction data record 1151, a total of 60258.2 gene -- drug adverse reaction network correspondence, association rules algorithm is set to Lift2, there were 829 signals, Yule 's Q, PRR, ROR, 2, MRHA positive signal coincidence rate were 73.95%, 57.39%, 57.39%, 24.65%, 3.81%.3, the algorithm of association rules mining results screened 4 genes -- adverse reactions associated signals strong drugs, are eylea, Wu Na paim monoclonal antibody, monoclonal antibody, cetuximab and.4. Screening of association rules algorithm and frequency method (Yule' s Q, PRR, ROR, 2) - gene mining results in the adverse reactions of signal strength and a high degree of coincidence of the paired with related analysis of drug adverse reactions, respectively study the various parts eylea bleeding, skin paim monoclonal antibody Study on the adverse reactions and adverse reactions of cetuximab in the respiratory system. The conclusion of the study: This study established a monoclonal antibody targeting antitumor drugs -- gene - adverse reaction relationship network, combined with the association rules mining algorithm and frequency method for signal comparison mining; association rule algorithm and frequency signal mining results, found two the method of coincidence degree is better; screened 4 genes -- strong signal drug adverse reaction incidence; through the literature search and database query, the high correlation of signal analysis, found that the signal mining results to further research have certain reference; this research belongs to the basic research, provides preliminary data reference and theoretical basis for the further research on drug epidemiology research and monoclonal antibody targeted antitumor drug adverse reaction.

【学位授予单位】：第二军医大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R979.1

【参考文献】