Department of Endocrinology, Beijing Tongren Hospital, Capital Medical University, Beijing, China.
Department of Mathematics, School of Biomedical Engineering, Capital Medical University, Beijing, China.
Biomed Res Int. 2019 Jan 13;2019:9565794. doi: 10.1155/2019/9565794. eCollection 2019.
The pathogenesis Graves' Orbitopathy (GO) is not yet fully understood. Here, we conducted a pathway analysis based on genome-wide DNA methylation data of Chinese GO patients to explore GO-related pathways and potential feature genes.
Six GO patients and 6 age-matched control individuals were recruited, and a genome-scale screen of DNA methylation was measured using their peripheral blood sample. After extracting the differentially methylated regions (DMRs), we classified DMRs into three clusters with respect to median absolute deviation (MAD) for GO and control group, respectively. Then the extract tests were performed to identify significant pathways by comparing the counts of genes in each cluster between GO and control group in a pathway. For each significant pathway, we calculated the Methylation-based Inference of Regulatory Activity (MIRA) score to infer the regulatory activity of genes involved in the pathway. Furthermore, we took the significant pathways as the subsets and applied Random forests (RF) method to extract GO-related feature genes.
We identified four potential significant pathways associated with the occurrence and development of GO disease. There were Toxoplasmosis, Axon guidance, Focal adhesion, and Proteoglycans in cancer (p<0.001 or p=0.007). The identified genes involved in the significant pathways, such as LDLR (p=0.019), CDK5 (p=0.036), and PIK3CB (p=0.020), were found to be correlated with GO phenotype.
Our study suggested pathway analyses can help understand the potential relationships between the DNA methylation level of some certain genes and their regulation in Chinese GO patients.
Graves 眼病（GO）的发病机制尚未完全阐明。在此，我们对中国 GO 患者的全基因组 DNA 甲基化数据进行了通路分析，以探讨 GO 相关通路和潜在特征基因。
招募了 6 例 GO 患者和 6 名年龄匹配的对照者，使用其外周血样本进行全基因组 DNA 甲基化筛选。提取差异甲基化区（DMRs）后，我们分别根据 GO 和对照组中位数绝对偏差（MAD）将 DMRs 分为三个簇。然后通过比较 GO 和对照组每个簇中基因的数量，对提取的检验进行GO 和对照组之间的通路进行显著通路的识别。对于每个显著通路，我们计算了基于甲基化的调控活性推断（MIRA）评分，以推断通路中涉及的基因的调控活性。此外，我们将显著通路作为子集，并应用随机森林（RF）方法提取 GO 相关特征基因。
我们确定了四个与 GO 疾病发生和发展相关的潜在显著通路。分别是弓形体病、轴突导向、黏着斑和癌症中的蛋白聚糖（p<0.001 或 p=0.007）。在显著通路中鉴定的基因，如 LDLR（p=0.019）、CDK5（p=0.036）和 PIK3CB（p=0.020），与 GO 表型相关。
我们的研究表明，通路分析有助于了解某些特定基因的 DNA 甲基化水平与其在中国 GO 患者中的调控之间的潜在关系。
