基于生物信息学分析双硫死亡相关基因在代谢相关脂肪性肝病中的作用  

Mechanistic study on the role of disulfidptosis-related genesin metabolism-associated fatty liver disease

作　　者：熊永强 王博[1] 王纪云 李韧[1] 张澍[1,2] XIONG Yongqiang;WANG Bo;WANG Jiyun;LI Ren;ZHANG Shu(Department of Geriatric Surgery,The Second Affiliated Hospital of Xi’an Jiaotong University,Xi’an 710004;Experimental Teaching Center for Clinical Skills,The Second Affiliated Hospital of Xi’an Jiaotong University,Xi’an 710004,China)

机构地区：[1]西安交通大学第二附属医院老年普通外科,陕西西安710004 [2]西安交通大学第二附属医院临床技能中心,陕西西安710004

出　　处：《西安交通大学学报(医学版)》2025年第2期249-256,共8页Journal of Xi’an Jiaotong University(Medical Sciences)

基　　金：陕西省重点研发计划一般项目(社会发展领域)(No.2020SF-072,2017JM8153);西安交通大学第二附属医院自由探索项目[No.2020YJ(ZYTS)018];西安交通大学医学“基础-临床”融合创新项目(No.YXJLRH2022062)。

摘　　要：目的基于生物信息学探讨双硫死亡相关基因(DRGs)在代谢相关脂肪性肝病(MAFLD)病情进展中的作用及相关机制。方法利用GEO数据库,筛选符合条件的MAFLD表达相关数据,进行差异基因分析,通过一致聚类识别DRGs并对MAFLD患者进行亚型分型。进一步评估亚型间的免疫浸润状态,运用CIBERSORT算法分析免疫细胞浸润情况。通过加权基因共表达网络分析(WGCNA)筛选与疾病相关的基因模块。然后,利用机器学习模型,基于DRGs筛选特征基因构建诊断模型,并对模型性能进行验证。最后,通过外部数据集评估DRGs在不同亚型间的稳定性,使用统计检验分析结果的显著性差异。结果通过分析数据集GSE31803,筛选出与MAFLD临床特征密切相关的DRGs共6个(SLC3A2、NCKAP1、CYFIP1、FLNA、MYL6、MYH10);将MAFLD患者分为两种亚型,亚型1具有更高的免疫细胞浸润水平;WGCNA识别出关键基因模块;通过机器学习筛选,支持向量机(SVM)模型被确定为最佳分类模型。外部验证确认了关键基因在MAFLD不同亚型中的稳定性和有效性。结论基于DRGs鉴定出两种高度异质性的MAFLD亚型,其临床特征、生物学过程和免疫状态均有显著差异,表明DRGs在MAFLD发生发展中起重要作用。Objective To explore the mechanism underlying the role of disulfidptosis-related genes(DRGs)in the disease progression of metabolically associated fatty liver disease(MAFLD)based on bioinformatics.Methods In this study,the GEO database was utilized to screen for eligible MAFLD expression data,conduct differential gene analysis,and identify DRGs through consistent clustering to subtype MAFLD patients.The immune infiltration status among subtypes was further evaluated,and the infiltration of immune cells was analyzed using the CIBERSORT algorithm.The gene modules related to the disease were selected through weighted gene co-expression network analysis(WGCNA).Subsequently,a diagnostic model was constructed based on DRGs using machine learning models,and the performance of the model was verified.Finally,the stability of DRGs among different subtypes was evaluated using an external dataset,and the significance of the results was analyzed using statistical tests.Results Through the analysis of the dataset GSE31803,six disulfide death genes,namely,SLC3A2,NCKAP1,CYFIP1,FLNA,MYL6 and MYH10,which were closely related to the clinical characteristics of MAFLD,were screened out.MAFLD patients were classified into two subtypes,with subtype 1 having a higher level of immune cell infiltration.Key gene modules were identified through WGCNA.Through machine learning screening,the support vector machine(SVM)model was determined as the optimal classification model.External validation confirmed the stability and effectiveness of the key genes in different subtypes of MAFLD.Conclusion Based on DRGs,two highly heterogeneous subtypes of MAFLD were identified,which exhibited significant differences in clinical characteristics,biological processes and immune status,indicating that DRGs play a crucial role in the occurrence and development of MAFLD.

关 键 词：代谢相关脂肪性肝病(MAFLD) 双硫死亡 免疫状态 生物信息学 

分 类 号：R575.29[医药卫生—消化系统]