烧伤疼痛是影响烧伤患者康复及生活质量的关键因素,其机制复杂,涉及多种炎症与免疫信号通路。本研究整合生物信息学和实验验证,旨在筛选鉴定与烧伤疼痛密切相关的关键基因,探索其作为潜在治疗靶点的临床价值。通过GEO数据库获取烧伤患者与健康对照的表达谱数据(GSE19743作训练集,GSE37069作验证集),结合GeneCaRNA数据库筛选疼痛相关基因。运用差异分析、GO与KEGG富集分析、构建可视化PPI网络,以及LASSO回归、SVM和RF三种机器学习方法识别关键基因,构建诊断模型,利用ROC曲线与DCA评估诊断效能。最终经RT-qPCR对外周血样本中候选基因表达水平进行实验验证。本研究筛选出117个烧伤疼痛相关差异表达基因,富集于PI3K-Akt、MAPK等炎症信号通路,IFNG、IL10和TLR4被三种机器学习方法共同识别为关键特征基因。基于此三基因构建的诊断模型在GSE37069验证集中表现优异,AUC达0.959。RT-qPCR验证表明,烧伤患者中IL10显著上调,IFNG表达下降,TLR4表达无显著差异,部分结果与生物信息学分析一致。IFNG、IL10和TLR4或通过调控免疫与炎症反应参与烧伤疼痛的发生维持,有望成为诊断生物标志物及治疗靶点,未来研究需进一步探讨其信号通路机制及临床干预价值。Burn pain is a critical factor affecting the recovery and quality of life of burn patients. Its underlying mechanisms are complex, involving multiple inflammatory and immune signaling pathways. This study integrates bioinformatics and experimental validation to identify key genes closely associated with burn pain and to explore their potential clinical value as therapeutic targets. Gene expression profiles (GSE19743 as the training set and GSE37069 as the validation set) were retrieved from the GEO database, and pain-related genes were screened via the GeneCaRNA database. Subsequent analyses included differential analysis, GO and KEGG enrichment analyses, and PPI network construction and visualization. Three machine learning algorithms—LASSO regression, SVM, and RF—were employed to identify key genes, following which a diagnostic model was established and evaluated using ROC curves and DCA. RT-qPCR validated candidate gene expression in peripheral blood samples. A total of 117 differentially expressed burn pain-related genes were identified, primarily enriched in inflammatory signaling pathways such as PI3K-Akt and MAPK. IFNG, IL10, and TLR4 were consistently identified as key feature genes across all three machine-learning methods. The diagnostic model based on these genes demonstrated excellent performance in the GSE37069 validation set, achieving an AUC of 0.959. RT-qPCR validation indicated that IL10 was significant
暂无评论