【摘要】 目的：通过生物信息学技术识别皮肤老化相关的作用靶基因。方法：通过合并转录组GSE85358和GSE67098信息分析和筛选皮肤老化相关的差异表达基因（DEGs）；进行蛋白质-蛋白质相互作用网络和功能分析，分别确定与DEGs相关的生物学功能和通路机制等。利用随机森林（RF）和人工神经网络（ANN）方法构建皮肤老化区分模型；筛选得到皮肤老化相关模型基因（MGs）后，基于GSE18876验证集，采用受试者工作特征曲线（ROC）验证其识别效能；利用免疫细胞亚型分布评估方法对皮肤老化相关免疫细胞的浸润情况进行分析；通过泛癌分析探索MGs的分析广度。结果：筛选获得了67个皮肤老化相关差异基因（AS-DEGs）。功能富集分析结果显示，67个DEGs主要参与机体疾病或机体失衡状态相关的机制通路，以及一些细胞结构和信号转导的生物学过程。通过构建RF和ANN模型筛选出4个重要的皮肤老化相关MGs(SOD3、LHFPL4、NEFH和LRG1)，训练集（GSE85358和GSE67098）和验证集（GSE18876）的模型曲线下面积（AUC）值分别为0.973和0.662。结论：与皮肤老化相关的4个基因，即SOD3、LHFPL4、NEFH和LRG1可能成为标志物用作皮肤老化相关的潜在药物靶点。更多还原

【Abstract】 Objective: To identify target genes related to skin aging through bioinformatics technology. Methods: Differential expression genes(DEGs) related to skin aging were screened through transcriptome information from GEO datasets(GSE85358 and GSE670988). A protein-protein interaction network was conducted; and a functional analysis was performed to determine the biological functions and underlying mechanisms associated with DEGs. A skin aging discrimination model was constructed using random forest(RF) and artificial neural network(ANN) methods. After screening for skin aging related model genes(MGs), ROC curve analysis was performed based on the GSE18876 validation set to verify their performance. The infiltration of immune cells related to skin aging was evaluated by assessing the distribution of various immune cell subtypes. The breadth of MGs analysis was then explored through a pan cancer analysis. Results: Sixty-seven differential expression genes(DEGs)related to skin aging were obtained through screening. The results of functional enrichment analysis showed that 67 GEGs were mainly involved in mechanism pathways related to body diseases or imbalanced states. Some biological processes were related to cellular structure and signal transduction. Four important skin aging related MGs(SOD3, LHFPL4, NEFH, and LRG1) were selected by constructing RF analysis and ANN models. The AUC validation values for the model in the training group(GSE85358 and GSE67098) and the testing group(GSE18876) were 0.973 and 0.662, respectively. Conclusion: Four genes,SOD3, LHFPL4, NEFH, and LRG1, are associated with skin aging; they could serve as potential drug targets and biomarkers for skin aging.更多还原