基于机器学习算法鉴定哮喘的坏死性凋亡相关生物标志物

Identification of necroptosis-related biomarkers in asthma based on machine learning algorithms

目的通过生物信息学分析的方法识别和验证哮喘中潜在的坏死性凋亡相关基因(necroptosis-related genes,NRGs)。方法基因表达综合(Gene Expression Omnibus,GEO)数据库提供了基因表达谱数据集GSE76262,并使用R软件筛选潜在的差异表达NRGs。对差异表达的NRGs进行蛋白质-蛋白质相互作用(protein-protein interaction,PPI)分析、基因本体论(gene ontology,GO)富集分析、京都基因与基因组百科全书(Kyoto encyclopedia of genes and genomes,KEGG)通路富集分析。通过最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)逻辑回归和支持向量机递归特征消除(support vector machine-recursive feature elimination,SVM-RFE)算法分析上调的NRGs。共同特征基因被确定为潜在的诊断标志物,并绘制受试者特征(receiver operator characteristic,ROC)曲线以验证其诊断效能。GSE137268进一步验证筛选出的特征基因的表达情况和诊断效能。使用在线工具预测可以靶向调控特征基因表达的微小RNAs(microRNAs,miRNAs)。结果在118例哮喘患者和21名健康对照者中鉴定出33个差异表达的NRGs(13个上调和20个下调)。PPI结果提示20个差异表达的NRGs相互作用。GO和KEGG富集分析显示NRGs与多个信号通路、淋巴细胞激活、细胞凋亡和免疫调节等相关。LASSO和SVM-RFE筛选出7个上调的NRGs可作为潜在的诊断基因,其ROC曲线显示出较高的诊断效率,曲线下面积(area under the curve,AUC)高于0.7。经GSE137268验证,7个特征基因与训练集的表达趋势相同(AUC>0.65)。预测hsa-miR-138-5p、hsa-miR-200b-3p和hsa-miR-30e-5p可调控哮喘患者NRGs的表达。结论BIRC3、HIF1A、FLOT1、NLRP3、RIPK2、GBE1和PELI1为哮喘的潜在生物标志物。Hsa-miR-138-5p、hsa-miR-200b-3p和hsa-miR-30e-5p分别是HIF1A、RIPK2和PELI1的上游调节因子。

Objective To identificate and validate of potential necroptosis-related genes(NRGs)in asthma through bioinformatics analysis.Methods The Gene Expression Omnibus(GEO)database provided the gene expression profile dataset GSE76262,and R software was used to screen for potential differentially expressed NRGs.Protein-protein interactions(PPI)analysis,gene ontology(GO)enrichment analysis,and Kyoto encyclopedia of genes and genomes(KEGG)pathway enrichment analysis were performed on the differentially expressed NRGs.The up-regulated NRGs were analyzed by least absolute shrinkage and selection operator(LASSO)logistic regression and support vector machine-recursive feature elimination(SVM-RFE)algorithms.Common signature genes were identified as potential diagnostic markers,and receiver operator characteristic(ROC)curves were drawn to verify their diagnostic efficacy.The expression and diagnostic efficacy of the screened signature genes were further confirmed by GSE137268.Online tools were used to predict microRNAs(miRNAs)that can target and regulate the expression of signature genes.Results Thirty-three differentially expressed NRGs(13 up-regulated and 20 down-regulated)were identified in 118 asthma patients and 21 healthy controls.The PPI results indicated that 20 differentially expressed NRGs interacting.GO and KEGG enrichment analyses revealed enrichment items related to multiple signaling pathways,lymphocyte activation,apoptosis,and immune regulation.LASSO and SVM-RFE showed that seven up-regulated NRGs could be potential diagnosis genes.The ROC curves showed high diagnostic efficiency with the area under the curve(AUC)higher than 0.7.GSE137268 verified that seven signature genes showed the same expression trend as the training set(AUC>0.65).Hsa-miR-138-5p,hsa-miR-200b-3p and hsa-miR-30e-5p were predicted to regulate the expression of necroptosis genes in asthmatic patients.Conclusion BIRC3,HIF1A,FLOT1,NLRP3,RIPK2,GBE1 and PELI1 can serve as potential biomarkers for asthma.Hsa-miR-138-5p,hsa-miR-200b-3p and hsa-miR-30e-5p are upstream regulators of HIF1A,RIPK2 and PELI1,respectively.