期刊文献+

基因标签区分癌症样本的初步分析

Preliminary Analysis of Cancer Sample Classification Based on Gene Signatures
原文传递
导出
摘要 从基因表达水平的角度对癌症样本进行分子分型,其较高的临床结果预测能力已经获得越来越广泛的关注。研究发现,某些基因集合(称为基因标签)的表达水平可以显著区分不同癌症样本之间的亚型差异并与癌症病人的临床结果显著关联。使用主成分分析法能够帮助我们找出这类表达谱与临床结果密切相关的基因标签。本研究通过主成分分析法从MSig BD和GeneSigDB数据库包含的基因集中找出可以解释不同癌症样本区别的基因标签并进一步评价了这些基因标签对癌症病人生存结果的预测能力。通过对10种癌症分别进行检验,我们找到了数量可观的可以区分癌症样本且相应的临床结果具有显著差异的基因标签。基于基因表达谱、临床数据和主成分分析法,基因标签能够对具有不同临床结果的癌症样本进行亚类的区分。 Molecular classification of cancer samples according to gene expression levels has attracted more and more attention because of its good predictive power for clinical outcomes. The expression profiles of some gene sets, named gene signatures, are sufficient for classifying cancer samples into subtypes associated with different clinical outcomes. It will be very useful to identify gene signatures whose gene expression profiles are predictive of clinical outcomes based on principal component analysis. In this study, we had used principal component analysis to identify gene signatures from MSigDB and GeneSigDB databases to explain differences of cancer samples and further assessed their power on predicting patients' survival outcomes. For 10 types of cancers, we had identified many gene signatures that are able to distinguish cancer samples and associate them with different clinical outcomes. Based on principal component analysis, gene signatures can be used to identify cancer subtypes with significantly different survival outcomes.
作者 刘秋芳 孔华磊 程酩 Liu Qiufang;Kong Hualei;Cheng Ming(School of Biomedical Engineering,Shanghai Jiao Tong University,Shanghai,200240;Shanghai Center for Systems Biomedicine,Shanghai Jiao Tong University,Shanghai,200240)
出处 《基因组学与应用生物学》 CAS CSCD 北大核心 2018年第10期4682-4688,共7页 Genomics and Applied Biology
基金 国家自然科学基金(21273148) 国家重大仪器专项(2014YQ090709) 上海市科学技术委员会(15142201200) 上海交通大学“医工(理)交叉研究基金”(YG2012MS58,YG2016QN66)共同资助
关键词 基因标签 癌症分类 生存结果 Gene signature Cancer classification Survival outcomes
  • 引文网络
  • 相关文献
;
使用帮助 返回顶部