引用本文:徐 畅,王 雪,郭 鑫,李 毅,赵 船 ,侯跃芳.基于疾病数据库的文本挖掘工具对比研究[J].中华医学图书情报杂志,2018,27(6):10-15.
基于疾病数据库的文本挖掘工具对比研究
Disease databases-based comparison of text mining tools
DOI:10.3969/j.issn.1671-3982.2018.06.002
中文关键词:  疾病数据库  文本挖掘工具  疾病候选基因  比较研究  实证研究
英文关键词:Disease database  Text mining tool  Candidate gene of disease  Compareilive stucly  Empirical study
基金项目:辽宁省教育厅人文社会科学研究项目“基于复杂网络的非相关文献知识发现方法研究”(LR201606);2017年辽宁省大学生创新创业训练项目“基于疾病遗传数据库的文本挖掘研究”(201710159000031)
作者单位
徐 畅 中国医科大学医学信息学院辽宁 沈阳 110122 
王 雪  
郭 鑫  
李 毅  
赵 船   
侯跃芳  
摘要点击次数: 98
全文下载次数: 17
中文摘要:
      目的:对比分析几种基于疾病数据库的文本挖掘工具,总结各自特点。 方法:选择eRAM、PhenUMA、Gendoo、G2D 4种工具进行对比分析。以Rett综合征为例进行实证研究,筛选与其关联性较高的基因,并结合先验知识做出预测。 结果:eRAM、PhenUMA知识库功能全面,可视化效果好。通过实证研究得到Rett综合征相关基因,并结合PubMed、UniProt等数据库中的先验知识推测出基因EGR2、CDKL5、BCHE、DLX5与Rett综合征相关。结论:基于疾病数据库的文本挖掘工具可以有效预测疾病的相关基因,预测疾病、表型、基因间相似和相关关系,有助于疾病的诊断及治疗。
英文摘要:
      Objective To comparatively analyze the disease databases-based text mining tools and their characteristics. Methods The 4 text mining tools, namely eRAM, PhenUMA, Gendoo and G2D4, were comparatively analyzed. The high correlation genes in Rett syndrome were screened and predicted according to the priori knowledge in PubMed and UniProt. Results The functions and visualization effect of eRAM and PhenUMA were good. The Rett syndrome-related genes were detected in the empirical study which showed that the EGR2, CDK15, BCHF and DLX5 genes were related with Rett syndrome according to the priori knowledge in PubMed and UniProt. Conclusion Disease databases-based text mining tools can effectively predict disease-related genes, similarity and correlation among diseases, phenotypes and genes, and can thus contribute to the diagnosis and treatment of diseases.
查看全文  查看/发表评论  下载PDF阅读器
关闭
手机扫一扫看