加权ceRNA网络筛选乳腺癌生物标志物

Screening of Breast Cancer Biomarkers by Weighted ceRNA Network

竞争性内源RNA(competing endogenous RNA,ceRNA)作为生物标志物和潜在的治疗靶点,在探究肿瘤的发病机制中,表现出了巨大的研究价值和临床应用前景。本文对乳腺癌ceRNA网络进行了系统的分析,首先通过差异分析获得差异miRNA、mRNA和lncRNA,其次利用网络的边聚集系数(edge clustering coefficient,ECC)和皮尔逊相关系数(Pearson correlation coefficient,PCC)计算ceRNA网络节点的权重,最后采用基于随机森林的逐步特征选择(stepwise feature selection based on random forest,SFS-RF)方法筛选出一组可作为乳腺癌生物标志物的RNA———LINC00466、CHL1-AS2和LINC00337,并利用GEO数据库验证了该组RNA对乳腺癌样本的识别情况。此外,通过GO和KEGG通路富集分析探索了该组RNA在乳腺癌中的生物学功能。结果显示:这些RNA作为生物标志物,在识别乳腺癌样本方面具有高精度和高效率等特性,在乳腺肿瘤细胞的增殖及遗传物质的合成等过程中具有重要生物学意义。

As a biomarker and a potential therapeutic target,competing endogenous RNAs(ceRNAs)have shown great research values and clinical application prospects in exploring the pathogenesis of tumors.Herein,the breast cancer ceRNA network was systematically analyzed.Firstly,differential RNAs were obtained by differential analysis.Secondly,the weight of the ceRNA network node was calculated by the edge clustering coefficient(ECC)and the Pearson correlation coefficient(PCC).Finally,LINC00466,CHL1-AS2 and LINC00337 were selected as biomarkers by the stepwise feature selection based on random forest(SFS-RF),and were used to identify breast cancer samples in GEO database.In addition,the biological function of this group of RNAs in breast cancer was explored by GO and KEGG pathway enrichment analysis.The results showed that this group of RNAs,used as a biomarker,has high accuracy and high efficiency in identifying breast cancer samples,and has important biological significance in the proliferation of breast cancer cells and the synthesis of genetic material.