检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:危思思 唐小凯 潘亭玉 雷恩骏[1] WEI Si-si;TANG Xiao-kai;PAN Ting-yu;LEI En-jun(Department of Anesthesiology,the First Affiliated Hospital of Nanchang University,Nanchang 330006,China;Department of Orthopedics,the First Affiliated Hospital of Nanchang University,Nanchang 330006,China)
机构地区:[1]南昌大学第一附属医院麻醉科,南昌330006 [2]南昌大学第一附属医院骨科,南昌330006
出 处:《南昌大学学报(医学版)》2024年第4期29-34,58,共7页Journal of Nanchang University:Medical Sciences
基 金:江西省研究生创新专项资金项目(YC2022—s085)。
摘 要:目的利用机器学习模型筛选出糖尿病肾病(DN)的诊断标志基因。方法从GEO数据库中获取4个与DN相关的转录组数据集,使用R语言对从数据集进行整合,消除批次效应。利用“limma”包筛选出健康对照组和DN组之间的差异表达基因(DEGs)。使用LASSO回归、随机森林(RF)和支持向量机(SVM)3种机器学习模型分别筛选与DN相关性最显著的特征基因,并利用“Venn”包对不同算法筛选的DEGs与铁死亡相关基因取交集,获得与铁死亡相关的DN诊断标志基因。通过差异分析、ROC曲线分析、体内和体外实验验证诊断标志基因的差异表达和诊断效能。结果共筛选出80个DEGs,机器学习模型筛选的结果和铁死亡相关基因的交集共得出3个诊断标志基因,其中锌指蛋白36(ZFP36)在训练组和验证组中都存在表达差异,且曲线下面积(AUC)>0.7。小鼠DN模型免疫组织化学和细胞蛋白免疫印迹的实验结果均表明,ZFP36在DN组中表达量显著降低(P<0.05)。结论ZFP36对DN具有良好的诊断效能,有望成为与铁死亡相关的DN诊断标志物及治疗靶点。Objective To screen out diagnostic marker genes for diabetic nephropathy(DN)using machine learning models.Methods The microarray dataset downloaded from the GEO database was integrated using R language and the batch effect was eliminated.The“limma”package was used to screen for the differentially expressed genes(DEGs)between the healthy control group and the DN group.3 machine learning models,namely,LASSO regression,Random Forest(RF)and Support Vector Machine(SVM),were used to screen out marker genes with the most significant correlation with DN,and the intersection of the DEGs screened by different algorithms and ferroptosis-related genes was performed using the“Venn”package,to obtain the ferroptosis-related diagnostic markers of DN.Differential expression and diagnostic efficacy of diagnostic marker genes were verified by differential analysis,ROC curve analysis,in vivo and in vitro experiments.Results A total of 80 DEGs were screened,and the intersection of the results of machine learning models and ferroptosis-related genes yielded a total of three diagnostic marker genes,among which zinc finger protein 36(ZFP36)was differentially expressed in both the training and validation groups with an area under the curve(AUC)of>0.7.The experimental results of both immunohistochemistry and cellular protein immunoblotting in the mouse DN model showed that the expression of ZFP36 was significantly reduced in the DN group(P<0.05).Conclusions ZFP36 can have potential diagnostic efficacy for DN and can be idealized as a ferroptosis-related diagnostic marker gene and therapeutic target for DN.
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.195