Liao Shenling, He He, Zeng Yuping, Yang Lidan, Liu Zhi, An Zhenmei, Zhang Mei
Department of Laboratory Medicine, West China Hospital, Sichuan University, Chengdu 610041, Sichuan, China.
Department of Endocrinology and Metabolism, West China Hospital, Sichuan University, Chengdu, China.
Open Med (Wars). 2021 May 17;16(1):773-785. doi: 10.1515/med-2021-0286. eCollection 2021.
To identify differentially expressed and clinically significant mRNAs and construct a potential prediction model for metabolic steatohepatitis (MASH).
We downloaded four microarray datasets, GSE89632, GSE24807, GSE63067, and GSE48452, from the Gene Expression Omnibus database. The differentially expressed genes (DEGs) analysis and weighted gene co-expression network analysis were performed to screen significant genes. Finally, we constructed a nomogram of six hub genes in predicting MASH and assessed it through receiver operating characteristic (ROC) curve, calibration plot, and decision curve analysis (DCA). In addition, qRT-PCR was used for relative quantitative detection of RNA in QSG-7011 cells to further verify the expression of the selected mRNA in fatty liver cells.
Based on common DEGs and brown and yellow modules, seven hub genes were identified, which were NAMPT, PHLDA1, RALGDS, GADD45B, FOSL2, RTP3, and RASD1. After logistic regression analysis, six hub genes were used to establish the nomogram, which were NAMPT, RALGDS, GADD45B, FOSL2, RTP3, and RASD1. The area under the ROC of the nomogram was 0.897. The DCA showed that when the threshold probability of MASH was 0-0.8, the prediction model was valuable to GSE48452. In QSG-7011 fatty liver model cells, the relative expression levels of NAMPT, GADD45B, FOSL2, RTP3, RASD1 and RALGDS were lower than the control group.
We identified seven hub genes NAMPT, PHLDA1, RALGDS, GADD45B, FOSL2, RTP3, and RASD1. The nomogram showed good performance in the prediction of MASH and it had clinical utility in distinguishing MASH from simple steatosis.
鉴定差异表达且具有临床意义的mRNA，并构建代谢性脂肪性肝炎（MASH）的潜在预测模型。
我们从基因表达综合数据库下载了四个微阵列数据集，即GSE89632、GSE24807、GSE63067和GSE48452。进行差异表达基因（DEG）分析和加权基因共表达网络分析以筛选重要基因。最后，我们构建了一个由六个枢纽基因组成的预测MASH的列线图，并通过受试者工作特征（ROC）曲线、校准图和决策曲线分析（DCA）对其进行评估。此外，使用qRT-PCR对QSG-7011细胞中的RNA进行相对定量检测，以进一步验证所选mRNA在脂肪肝细胞中的表达。
基于共同的DEG以及棕色和黄色模块，鉴定出七个枢纽基因，分别为NAMPT、PHLDA1、RALGDS、GADD45B、FOSL2、RTP3和RASD1。经过逻辑回归分析后，使用六个枢纽基因建立列线图，分别为NAMPT、RALGDS GADD45B、FOSL2、RTP3和RASD1。列线图的ROC曲线下面积为0.897。DCA显示，当MASH的阈值概率为0-0.8时，预测模型对GSE48452有价值。在QSG-7011脂肪肝模型细胞中，NAMPT、GADD45B、FOSL2、RTP3、RASD1和RALGDS的相对表达水平低于对照组。
我们鉴定出七个枢纽基因NAMPT、PHLDA1、RALGDS、GADD45B、FOSL2、RTP3和RASD1。该列线图在预测MASH方面表现良好，在区分MASH与单纯性脂肪变性方面具有临床实用性。
