請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/19909
標題: | 基因肥胖交互作用之於空腹血糖與糖化血色素的影響 Gene-obesity interactions on fasting glucose and glycated hemoglobin |
作者: | Yu-Ting Huang 黃鈺婷 |
指導教授: | 林菀俞(Wan-Yu Lin) |
關鍵字: | 糖尿病,基因環境交互作用,基因風險分數,腰臀比,臺灣人體生物資料庫, diabetes,gene-by-environment interactions,genetic risk score,waist-hip ratio,Taiwan Biobank, |
出版年 : | 2020 |
學位: | 碩士 |
摘要: | 臺灣糖尿病的盛行率約為6.45%。全世界有多項研究調查了遺傳和環境因素對糖尿病的影響。然而,在臺灣很少有研究探討遺傳對表型的影響是否會因抽菸或肥胖等因子的影響而變。本研究包括來自臺灣人體生物資料庫18,425名受試者,每位受試者具有587,526個單核苷酸多態性(SNPs)以及血液和尿液檢查的結果。我們研究了兩個常用於診斷糖尿病的指標:空腹血糖(FG)和糖化血色素(HbA1c)。在我們的分析中,飲酒、規律運動、抽菸、腰臀比(WHR)和腰圍(WC)依序作為環境因素,而根據不同的表型以及環境因子,樣本數的範圍從18,381至18,383位參與者納入分析。年齡、性別、教育程度以及前10個基因主要成分作為調整因子。我們的目標是在SNP層級和基因層級的支持下,共定位在FG和HbA1c上與環境因子有交互作用的基因。在SNP層級,進行了SNP單點分析;在基因層級,我們不僅使用了六種基因分析方法,也先藉由PrediXcan和基因型組織表達文庫(GTEx)預測基因表現量,並使用基因表現量產生轉錄本風險分數(Transcriptome Risk Score, TRS),測試TRS與環境因子之間的交互作用,最後使用四個Z分數進行檢定。在整個研究中,SNP單點分析與TRS分析的顯著閾值分別設置為1×〖10〗^(-6)和5×〖10〗^(-5);在後續的分析中,六種基因分析與四個Z分數的顯著閾值分別設置為8.3×〖10〗^(-6)與1.25×〖10〗^(-5)。我們發現11個基因在SNP層級和基因層級上,與WC和WHR有交互作用,其中有3個基因在先前已被報導與空腹血糖、發炎、肥胖、葡萄糖等有關,於複製研究中11個基因有3個得到驗證,分別是:VPS54 subunit of GARP complex (VPS54)、alpha 1,4-galactosyltransferase (P blood group) (A4GALT)與cytochrome b5 reductase 3 (CYB5R3)。在發現研究中,Set Based gene EnviRonment InterAction (GxE) test和Ridge GxE檢驗支持了大多數基因與環境交互作用的存在;於TRS分析中,交互作用得到了不同組織的支持,最重要的是脂肪組織,高WHR增加了A4GALT皮下脂肪組織表現量對於空腹血糖的影響(p =6.74×〖10〗^(-8))。本研究發現,後天因素 (WC和WHR) 是重要的,因為它們與空腹血糖與糖化血色素基因效應之增強有關,期望未來可以用相同的概念,以不同的表型或是人群之資料作為研究,了解更多基因與環境交互作用對於人體之影響,提供生物與醫療方面之輔助,減低疾病所帶來的負擔。 The prevalence of diabetes in Taiwan is about 6.45%. Some studies have investigated the influences of genetic and environmental factors on diabetes in the world. However, few studies have looked into whether the genetic influences on phenotypes can be modulated by some factors such as smoking and obesity in Taiwanese. This study included 18,425 subjects from the Taiwan Biobank, each with 587,526 single-nucleotide polymorphisms (SNPs) and results of blood and urine tests. We investigated two indicators commonly used to diagnose diabetes: fasting glucose (FG) and glycated hemoglobin (HbA1c). In our analysis, alcohol consumption, regular physical exercise, cigarette smoking, waist-hip ratio (WHR) and waist circumference (WC) were served as the environmental factors in order. According to different phenotypes and environmental factors, the sample size ranged from 18,381 to 18,383 participants included in the analysis. Age, sex, education attainment, and the first 10 genetic principal components were served as covariates. We aim to co-localize genes interacting with environmental factors on FG and HbA1c, supported by both the SNP level and the Gene level. From the SNP perspective, SNP-by-environment interaction analysis was performed. From the Gene perspective, not only six gene-based methods analyses were performed, but also gene expressions were first predicted by PrediXcan and the Genotype-Tissue Expression (GTEx) library, and we transformed gene expressions into Transcriptome Risk Score (TRS), and then the interactions between TRS and environmental factors were tested. Finally, we conducted four Z score tests. Throughout this study, the significance level for SNP-by-environment interaction analysis and TRS-by-environment interaction analysis were set at 1×〖10〗^(-6) and 5×〖10〗^(-5), respectively; in the subsequent analysis, the significance level for six gene-based methods analyses and the four Z score tests were set at 8.3×〖10〗^(-6) and 1.25×〖10〗^(-5), respectively. We found that 11 genes present interactions with WHR and WC in both the SNP and Gene level. Among the 11 genes, 3 genes have been reported to be associated with FG, inflammation, obesity and glucose. 3 out of 11 genes were replicated: VPS54 subunit of GARP complex (VPS54)、alpha 1,4-galactosyltransferase (P blood group) (A4GALT) and cytochrome b5 reductase 3 (CYB5R3). In the discovery study, most of gene-by-environment interactions (GxE) are supported by the set-based GxE test and Ridge GxE Test. ; in the TRS-by-environment interaction analysis, the interactions are supported by different tissues, and the most significant tissue is adipose subcutaneous and we found high WHR increases the effect of A4GALT expression on FG in the adipose subcutaneous ( p =6.74×〖10〗^(-8) ). This study found non-genetic factors (WHR and WC) are important because they are associated with an exacerbation of genetic effects on FG and HbA1c. We hope that the same concept can be used in the future, in different phenotypes or population data, to explore more GxE impacts on the human beings, to provide insights to the biological and medical fields, and to reduce the burden of diseases. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/19909 |
DOI: | 10.6342/NTU202003406 |
全文授權: | 未授權 |
顯示於系所單位: | 流行病學與預防醫學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
U0001-1408202013521900.pdf 目前未授權公開取用 | 2.23 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。