杜亚娟, 赵恩法, 张玉顺. 基于支持向量机递归特征消除筛选急性心肌梗死相关免疫基因诊断急性心肌梗死的价值[J]. 心脏杂志, 2020, 32(5): 471-475, 497. DOI: 10.12125/j.chj.202005029
    引用本文: 杜亚娟, 赵恩法, 张玉顺. 基于支持向量机递归特征消除筛选急性心肌梗死相关免疫基因诊断急性心肌梗死的价值[J]. 心脏杂志, 2020, 32(5): 471-475, 497. DOI: 10.12125/j.chj.202005029
    Ya-juan DU, En-fa ZHAO, Yu-shun ZHANG. Identification of immune-related genes in acute myocardial infarction based on support vector machine-recursive feature elimination[J]. Chinese Heart Journal, 2020, 32(5): 471-475, 497. DOI: 10.12125/j.chj.202005029
    Citation: Ya-juan DU, En-fa ZHAO, Yu-shun ZHANG. Identification of immune-related genes in acute myocardial infarction based on support vector machine-recursive feature elimination[J]. Chinese Heart Journal, 2020, 32(5): 471-475, 497. DOI: 10.12125/j.chj.202005029

    基于支持向量机递归特征消除筛选急性心肌梗死相关免疫基因诊断急性心肌梗死的价值

    Identification of immune-related genes in acute myocardial infarction based on support vector machine-recursive feature elimination

    • 摘要:
        目的  基于支持向量机(SVM)递归特征消除(RFE)(SVM-RFE)筛选急性心肌梗死(AMI)相关的关键免疫基因,并探讨其在AMI中的诊断价值。
        方法  通过基因芯片表达数据库(GEO)获取AMI患者基因表达芯片数据集GSE66360。通过SVM-RFE筛选AMI相关的免疫基因,进而用受试者工作特征(ROC)曲线评估其对AMI的诊断价值。
        结果  共得到66个差异免疫基因,基于SVM-RFE最终得到8个免疫基因用于构建SVM分类器,该基因分类器在探索队列(AUC=0.998)和验证队列(AUC=1.00)中都具有极好的预测AMI的能力。通路富集分析特征基因主要参与血管平滑肌细胞增殖的调控及血管动脉硬化等过程。
        结论  本研究识别出8个关键的免疫基因,发现其在AMI诊断中的潜在作用,并将增进对AMI发生的分子机制的了解。

       

      Abstract:
        AIM  To identify immune-related genes involved in acute myocardial infarction (AMI) based on support vector machine-recursive feature elimination and to further investigate their value in the diagnosis of AMI.
        METHODS  GSE66360 dataset of AMI gene expressions was downloaded from the Gene Expression Omnibus. Differentially expressed immune-related genes involved in AMI based on SVM-RFE were screened and their value in the diagnosis of AMI was further investigated. Receiver operating characteristic (ROC) was used to evaluate its diagnostic value for AMI.
        RESULTS  A total of 66 differentially expressed immune genes were obtained. Eight genes were finally identified based on SVM-RFE and they were used to construct a SVM classifier. The classifier had an excellent capacity of predicting AMI in both the discovery cohort (AUC=0.998) and the validation cohort (AUC=1.00). Pathways enrichment analysis revealed that the identified immune genes were mainly involved in the regulation of vascular smooth muscle cell proliferation and vascular arteriosclerosis.
        CONCLUSION  Our study has identified 8 immune genes and studied their potential roles in the diagnosis of AMI, which will improve our understanding of the molecular mechanism of the occurrence of AMI.

       

    /

    返回文章
    返回