Abstract: Background　Graves' Disease（GD） is a common autoimmune disease，but its pathogenesis is not clear.Objective　To find out genes that may related to the pathogenesis of GD by reanalyzing public repository for microarray gene data.Methods　"Graves' Disease" was searched in GEO database and ArrayExpress database，and GSE71956 and E-MEXP-2612 were chosen for further study.The "Limma" package in R language was used to standardize the original data of microarray and to identify the genes with expressions different from those of the normal people.The identified different expression genes were found in CD4 cells in controls compared with those of GD patients（C-GD CD4），in CD8 cells in controls compared with those of GD patients（C-GD CD8），in the thyroid tissue in controls compared with that of patients with short course GD（C-S），in the thyroid tissue in controls compared with that of patients with long course GD（C-L），and in the thyroid tissue in patients with short course GD compared with that of those with long course of GD（S-L）.Then，multiple comparisons and functional annotation analysis of the different expression genes were carried out.Results　Venn diagram showed that KLF9 and RGS1 existed in C-GD CD4，C-GD CD8 and C-L different expression genes simultaneously.Amino acid and steroid metabolism genes may be related to the pathogenesis of GD.FMO2，CALHM6 and C7 exist simultaneously in C-S，C-L and S-L different expression genes.Both long- and short-term GD patients showed higher mean expression level of CALHM6 in thyroid tissue compared with the control group（P<0.05）.The mean expression level of CALHM6 in long-term GD patients was higher than that of short-term GD patients（P<0.05）.Both long- and short-term GD patients demonstrated lower mean expression levels of FMO2 and C7 in thyroid tissue than the control group（P<0.05）.And the mean expression levels of FMO2 and C7 in long-term GD patients were lower than those of short-term GD patients（P<0.05）.Conclusion　The pathogenesis of GD may be related to the abnormal expression of KLF9 and RGS1 and the metabolism of amino acids and steroids.FMO2，CALHM6 and C7 may be related to the course of GD.

Key words: Graves disease, Thyroid gland, Gene microarray, Immune system, Steroid hormone

摘要： 背景　Graves' Disease（GD）是一种常见的自身免疫性疾病，但其发病机制尚不明确。目的　通过对公共基因芯片数据的分析，找出与GD发病可能相关的基因。方法　在GEO数据库和ArrayExpress数据库中检索“Graves' Disease”，得到GSE71956和E-MEXP-2612（截至2019-04-02）。利用R语言的“Limma”包对基因芯片的原始数据进行标准化处理并找出差异基因。差异基因共包括对照CD4细胞与GD患者CD4细胞的对比（C-GD CD4）、对照CD8细胞与GD患者CD8细胞的对比（C-GD CD8）、对照者甲状腺组织与短病程GD患者甲状腺组织的对比（C-S）、对照者甲状腺组织与长病程GD患者甲状腺组织的对比（C-L）、短病程GD患者甲状腺组织与长病程GD患者甲状腺组织的对比（S-L）。再对差异基因进行多重比较及功能注释分析。结果　维恩图显示，KLF9、RGS1基因同时存在于C-GD CD4、C-GD CD8与C-L差异基因中。氨基酸和类固醇代谢基因可能与GD的发病相关。FMO2、CALHM6和C7基因同时存在于C-S、C-L、S-L差异基因中。长病程和短病程GD患者甲状腺组织CALHM6基因表达水平高于对照者，长病程GD患者CALHM6基因表达水平高于短病程GD患者；长病程和短病程GD患者甲状腺组织FMO2和C7基因表达水平低于对照者，长病程GD患者FMO2和C7基因表达水平低于短病程GD患者（P<0.05）。结论　GD的发病可能与KLF9、RGS1的异常表达及氨基酸和类固醇代谢有关，FMO2、CALHM6、C7基因可能与GD的病程相关。

关键词: 格雷夫斯病, 甲状腺, 基因芯片, 免疫系统, 类固醇激素