背景:肥胖是一种复杂的疾病,受多种因素的影响,包括环境,行为,和遗传学。然而,肥胖的遗传机制尚不清楚.因此,我们的研究旨在使用生物信息学分析来确定人类肥胖的关键基因.
方法:从基因表达综合(GEO)数据库下载人类脂肪组织的微阵列数据集。在选择差异表达基因(DEGs)后,我们使用Lasso回归和支持向量机(SVM)算法进一步识别特征基因。此外,免疫细胞浸润分析,基因集变异分析(GSVA),进行了GeneCards数据库和转录调控分析,以研究特征基因可能影响肥胖的潜在机制。我们利用接收操作特征(ROC)曲线分析特征基因的诊断效能。最后,我们在细胞实验和动物实验中验证了特征基因。使用SPSS版本28.0进行验证实验中的统计分析,并使用GraphPadPrism9.0软件生成图形。使用R语言(版本4.2.2)进行生物信息学分析,使用显著性阈值p<0.05。
结果:使用Limma软件包选择了199个DEG,随后,5个特征基因(EGR2、NPY1R、通过Lasso回归和SVM算法选择GREM1,BMP3和COL8A1)。通过各种生物信息学分析,我们发现了一些信号通路,通过这些信号通路,特征基因影响肥胖,也揭示了这些基因在免疫微环境中的关键作用,以及它们与肥胖相关基因的强相关性。此外,ROC曲线显示所有特征基因对肥胖具有良好的预测和诊断效能。最后,经过体外实验验证,EGR2、NPY1R和GREM1被鉴定为关键基因。
结论:本研究确定EGR2、GREM1和NPY1R是人类肥胖的潜在关键基因和潜在诊断生物标志物。此外,EGR2首次被发现是人类脂肪组织中肥胖的关键基因,这可能为诊断和治疗肥胖症提供新的靶点。
BACKGROUND: Obesity is a complex condition that is affected by a variety of factors, including the environment, behavior, and genetics. However, the genetic mechanisms underlying obesity remains poorly elucidated. Therefore, our study aimed at identifying key genes for human obesity using bioinformatics analysis.
METHODS: The microarray datasets of adipose tissue in humans were downloaded from the Gene Expression Omnibus (GEO) database. After the selection of differentially expressed genes (DEGs), we used Lasso regression and Support Vector Machine (SVM) algorithm to further identify the feature genes. Moreover, immune cell infiltration analysis, gene set variation analysis (GSVA), GeneCards database and transcriptional regulation analysis were conducted to study the potential mechanisms by which the feature genes may impact obesity. We utilized receiver operating characteristic (ROC) curve to analysis the diagnostic efficacy of feature genes. Finally, we verified the feature genes in cell experiments and animal experiments. The statistical analyses in validation experiments were conducted using SPSS version 28.0, and the graph were generated using GraphPad Prism 9.0 software. The bioinformatics analyses were conducted using R language (version 4.2.2), with a significance threshold of p < 0.05 used.
RESULTS: 199 DEGs were selected using Limma package, and subsequently, 5 feature genes (
EGR2, NPY1R, GREM1, BMP3 and COL8A1) were selected through Lasso regression and SVM algorithm. Through various bioinformatics analyses, we found some signaling pathways by which feature genes influence obesity and also revealed the crucial role of these genes in the immune microenvironment, as well as their strong correlations with obesity-related genes. Additionally, ROC curve showed that all the feature genes had good predictive and diagnostic efficiency in obesity. Finally, after validation through in vitro experiments,
EGR2, NPY1R and GREM1 were identified as the key genes.
CONCLUSIONS: This study identified
EGR2, GREM1 and NPY1R as the potential key genes and potential diagnostic biomarkers for obesity in humans. Moreover,
EGR2 was discovered as a key gene for obesity in human adipose tissue for the first time, which may provide novel targets for diagnosing and treating obesity.