目的：基于microRNA（miRNA）调控网络预测卵巢癌多药耐药相关基因。方法： 综合运用文本挖掘、网络构建和预测等生物信息学分析方法，挖掘卵巢癌化疗耐药相关miRNA和miRNA-靶基因数据，并构建miRNA调控网络，利用已知miRNA对卵巢癌多药耐药相关基因进行预测。结果： 文本挖掘出11个与卵巢癌化疗耐药相关的miRNA，包括miR-130a、miR-214、let-7i、miR-125b、miR-376c、miR-199a、miR-93、miR-141、miR-130b、miR-193b*和miR-200c。在miRNA靶基因预测数据软件TargetScan中挖掘出47 077个miRNA-靶基因数据，而PicTar挖掘出1 675个miRNA-靶基因数据。在miRNA调控网络中，神经素1基因（neuropilins, NRP1)是最重要的Hub-基因。结论：利用已知miRNA构建miRNA调控网络进而预测卵巢癌多药耐药相关基因是一种行之有效的方法。NRP1极有可能在卵巢癌化疗耐药形成中扮演着重要的角色，是卵巢癌潜在的药物治疗靶点。

Objective: To predict genes related to multidrug resistance (MDR) in ovarian cancer based on the microRNA (miRNA) regulatory network. Methods: To identify potential genes associated with multidrug resistance in ovarian cancer based on published miRNAs and miRNA-target genes were identified using a comprehensive bioinformatics approaches including text mining and network researching. Results: Eleven miRNAs related to ovarian cancer chemoresistance were identified, including miR-130a, miR-214, let-7i, miR-125b, miR-376c, miR-199a, miR-93, miR-141, miR-130b, miR-193b* and miR-200c. A total of 47 077 putative targets were predicted using the TargetScan algorithm and 1 675 other targets were predicted using the PicTar algorithm. Neuropilins (NRP1) was the most important Hub-gene found in the cancer miRNA regulatory network. Conclusion: It is an effective method to construct a miRNA regulatory network to predict genes related to MDR in ovarian cancer by utilizing the existing information on miRNAs. Using this approach, we predict that NRP1 may play an important role in ovarian cancer chemoresistance and would be a potential therapeutic target for ovarian cancer.

国家高技术研究发展计划（863计划）资助项目（No. 2012AA02A507）；广西自然科学基金资助项目（桂财教2014-118号）；广西科学研究与技术开发计划课题资助项目（桂科攻14124004-1-24）。