基于机器学习方法筛选IgA肾病铁死亡关键基因

Screening key genes of ferroptosis in IgA nephropathy based on machine learning

目的本研究利用机器学习探索IgA肾病(IgAN)铁死亡的潜在诊断标志物及其生物学过程,并研究其与IgAN免疫细胞浸润的关系。方法从GEO数据库中下载来自IgAN患者和对照组的肾小球组织的微阵列数据集(GSE93798,GSE104948),通过FerrDb数据库获取全部铁死亡相关基因数据,获得IgAN组和对照组中有差异表达的铁死亡相关基因(P<0.05)。用LASSO回归、支持向量机和随机森林三种机器学习策略来确定潜在的IgAN铁死亡诊断标志物。为了评估这些潜在生物标志物的诊断效果,分别绘制了训练集和验证集的ROC曲线。此外,还采用CIBERSORT算法评估了IgAN肾组织的免疫细胞浸润情况,并研究了生物标志物和免疫细胞浸润之间的关系。结果共鉴定出157个IgAN组中的差异表达基因分析(DEGs);其中64个基因明显上调,93个基因明显下调。IgAN组和对照组中有特别显著差异表达的铁死亡相关基因52个。使用LASSO回归、支持向量机和随机森林三种机器学习策略最终确定了ZFP36被鉴定为IgAN的潜在诊断标志物。ZFP36在训练集的ROC曲线下面积为0.874,在验证集的ROC曲线下面积为0.874。免疫细胞浸润分析结果表明,静止树突细胞、中性粒细胞、单核细胞和静止的NK细胞可能参与了IgAN的发展。此外,ZFP36与这些免疫细胞亚型有不同程度的关联。结论铁死亡与IgAN的发生发展密切相关。ZFP36作为IgAN铁死亡的潜在诊断标志物,将为IgAN的诊断、治疗及预防提供新的靶点。

Objective To explore potential diagnostic markers and biological processes of ferroptosis in IgA nephropathy(IgAN),and to investigate its relationship with immune cell infiltration in IgAN using machine learning.Methods Microarray data sets(GSE93798,GSE104948)were downloaded from GEO database from IgAN patients and control groups.All ferroptosis related gene data were obtained by FerrDb database,and differential expression was obtained between IgAN group and control group of ferroptosis related genes(P<0.05).Three machine learning strategies,namely LASSO regression,support vector machine and random forest,were used to identify potential diagnostic markers of ferroptosis in IgAN.In order to evaluate the diagnostic effect of these potential biomarkers,ROC curves of the training set and the validation set were plotted respectively.In addition,CIBERSORT algorithm was used to evaluate the infiltration of immune cells in IgAN kidney tissue,and the relationship between biomarkers and immune cell infiltration was investigated.Results A total of 157 DEGs in IgAN groups were identified.Among them,64 genes were significantly up-regulated and 93 genes were significantly down-regulated.52 genes related to iron death were significantly differentially expressed between IgAN group and control group.Three machine learning strategies,namely LASSO regression,support vector machine and random forest,were used to identify ZFP36 as a potential diagnostic marker for IgAN.The area under ROC curve of ZFP36 was 0.874 in the training set and 0.874 in the verification set.Immune cell infiltration analysis indicated that stationary dendritic cells,neutrophils,monocytes,and stationary NK cells may be involved in the development of IgAN.In addition,ZFP36 was associated with these immune cell subtypes to varying degrees.Conclusion Ferroptosis is closely related to IgAN.ZFP36,as a potential diagnostic marker of ferroptosis in IgAN,will provide a new target for diagnosis,treatment and prevention of IgAN.