作者：杨启帆，方柔柔，邬东东，骆家引，徐守竹，赵晶

单位：

1.712046陕西省咸阳市，陕西中医药大学公共卫生学院　2.712046陕西省咸阳市，陕西中医药大学第二临床医学院　3.712046陕西省咸阳市，陕西中医药大学陕西省针药结合重点实验室

Units：

1.School of Public Health of Shaanxi University of Chinese Medicine, Xianyang 712046, China2.The Second Clinical Medical College of Shaanxi University of Chinese Medicine, Xianyang 712046, China3.Master Institute of Traditional Chinese Medicine of Shaanxi University of Chinese Medicine, Xianyang 712046, China

关键词：

糖尿病血管病变；糖尿病心脏病；铁死亡；生物信息学；LASSO回归

Keywords：

Diabetic angiopathies; Diabetic heart disease; Ferroptosis; Bioinformatics; LASSO regression

CLC：

R 587.23

DOI：

10.12114/j.issn.1008-5971.2023.00.271

Funds：

国家自然科学基金资助项目（82100488，82105016）；陕西省科技厅重点研发计划项目（2021SF-071，2022SF318）；陕西省教育厅重点科研计划项目（21JS012）；国家级大学生创新创业训练计划项目（202110716027）

摘要：

目的　基于生物信息学分析方法和机器学习算法探究铁死亡参与糖尿病心脏病（DHD）的机制。方法　从美国国家生物技术信息中心（NCBI）基因表达综合（GEO）数据库下载数据集GSE4745（大鼠心肌细胞测序数据）和数据集GSE26887（人心肌细胞测序数据），通过铁死亡数据库获得铁死亡基因数据集。应用R软件（4.2.1版本）中的“Limma”包分析数据集GSE4745中DHD的差异表达基因；应用“WGCNA”包构建加权基因共表达网络以筛选DHD相关基因模块；应用“Venn”包取DHD差异表达基因、DHD相关基因模块及铁死亡基因数据集交集以获取DHD相关铁死亡基因。然后针对DHD相关铁死亡基因进行LASSO回归分析以筛选DHD相关铁死亡核心基因。最后通过数据集GSE26887、蛋白质组学原始数据及既往研究验证DHD相关铁死亡核心基因。结果　从数据集GSE4745中获得491个DHD差异表达基因，其中上调基因252个、下调基因239个。以软阈值为7将DHD差异表达基因构建邻接矩阵，最终选择品红色和蓝色模块为DHD相关基因模块，共1 092个基因。进一步分析获得28个DHD相关铁死亡基因。LASSO回归分析结果显示，从28个DHD相关铁死亡基因中获得6个DHD相关铁死亡核心基因，分别为H19、Nr4a1、Decr1、Gstm1、Slc3a2、Por。根据数据集GSE26887、蛋白质组学原始数据及既往研究结果最终确定H19、Nr4a1、Decr1、Por为DHD相关铁死亡的核心基因。结论　铁死亡参与DHD的机制可能与H19、Decr1、Por表达上调及Nr4a1表达下调有关，这为铁死亡参与DHD提供了新的证据。

Abstract：

Objective To explore the mechanism of ferroptosis participating in diabetic heart disease (DHD) based onbioinformatics analysis methods and machine learning algorithms. Methods The dataset GSE4745 (rat cardiomyocyte sequencingdata) and dataset GSE26887 (human cardiomyocyte sequencing data) were downloaded from the Gene Expression Omnibus (GEO)database of National Center for Biotechnology Information (NCBI) . The ferroptosis gene dataset were selected from ferroptosisdatabase. The "limma" package in R software (version 4.2.1) was used to analyze the differentially expressed genes of DHD in thedataset GSE4745; the "WGCNA" package was used to construct a weighted gene coexpression network to screen DHD related genemodules; the "Venn" package was used to package the intersection of differentially expressed genes, DHD-related gene modulesand ferroptosis gene datasets, that DHD-related ferroptosis gene. Then LASSO regression analysis was performed on DHDrelated ferroptosis genes to screen DHD-related ferroptosis core genes. Finally, the DHD related ferroptosis core genes was verifiedby dataset GSE26887, proteomics raw data and previous studies. Results A total of 491 DHD differentially expressed geneswere obtained from the dataset GSE4745, including 252 up-regulated genes and 239 down regulated genes. DHD differentiallyexpressed genes were constructed into an adjacency matrix with a soft threshold of 7. Finally, magenta and blue modules wereselected as DHD related gene modules, with a total of 1 092 genes. Further analysis obtained DHD related ferroptosis genes. LASSO regression analysis results showed that six DHD related ferroptosis core genes were obtained from 28 DHD related ferroptosisgenes, which were H19, Nr4a1, Decr1, Gstm1, Slc3a2 and Por, respectively. H19, Nr4a1, Decr1 and Por were finally identified asthe DHD related ferroptosis core genes according to the dataset GSE26887, the proteomics raw data and previous research results.Conclusion The mechanism of ferroptosis participating in DHD may be related to the up regulation of H19, DECR1 and Porexpression and the down regulation of NR4A1 expression, which provides new evidence for ferroptosis participating in DHD.

ReferenceList：