請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/68937
標題: | 利用台灣生物資料庫探討成人肺功能下降之全基因體關聯研究 Genome-Wide Association Study of Pulmonary Function Decline in Taiwan Biobank Dataset |
作者: | Chia-Jung Lee 李佳容 |
指導教授: | 李永凌(Yungling Leo Lee) |
關鍵字: | 成人肺功能,成人肺功能預測公式,肺功能變化,單核甘酸多型性,全基因體關聯性分析, Adult pulmonary function,Pulmonary function predictive equation,Pulmonary function decline,Single nucleotide polymorphism,Genome-wide association study, |
出版年 : | 2016 |
學位: | 碩士 |
摘要: | 背景:肺功能檢測可以幫助了解目前肺部和氣道的生理狀況,提供許多臨床資料包括疾病診斷、嚴重度、治療成效、手術的可行性及危險性評估,使用肺量計(spirometry)得到之用力呼氣一秒量(forced expiratory volume in one second, FEV1)以及吐氣至最終的總吐氣量(forced vital capacity, FVC)是臨床上常使用的判斷依據,肺功能隨著年紀的成長而增加,在成人初期時達到頂峰,之後又隨著年紀的增加而逐年減少,抽菸以及空氣汙染等等的環境因子會加速肺功能的退化,遺傳因子也同樣影響肺功能逐年遞減的快慢。臨床上判斷肺功能時,需要以預測值為基準比較(%預測值),然而在台灣目前並沒有統一的肺功能預測公式。本研究利用台灣人體生物資料庫(Taiwan Biobank)的追蹤資料,建立台灣人肺功能的常模,以及進行全基因體關聯性研究尋找與成人肺功能下降有關的基因變異。
研究方法:本研究資料來源為台灣人體生物資料庫,主要為結合基因以及醫學資料進行台灣本土的世代追蹤計畫,自2012年正式開始收案,從台灣二十八個據點招募社區三十到七十歲且無癌症診斷的成年人。申請資料包含基因資料、肺功能檢測值、問卷調查之抽菸行為(曾經抽菸、二手菸暴露、抽菸)及疾病診斷(肺氣腫及氣喘)。研究分析分為兩個主要的部份,第一個部分為橫斷性設計,目的是建立並檢測台灣成人的肺功能預測常模,利用目前文獻上使用來建立與年齡變化相關的資料常模的兩個方法(Generalize additive models for location scale and shape and Quantile regression)在其中一部分資料進行模型建立,模型建立後,利用Mean Square Error, Mean Absolute Error, Akaike Information Criterion, Schwarz-Bayes Criterion等模型適配度檢測,在不同方法中進行模型選擇,並與2012年Global Lung Initiative和1997潘教授所建立的常模做比較,模型的驗證則利用極端值的預測以及內部交叉驗證(10-fold cross validation)。研究的第二部分,是使用世代追蹤的資料,計算預測肺功能變化以及實際測量之肺功能下降的差異值,先利用單變數分析進行相關因子的關聯性檢測,品質檢測後,利用基因資料進行全基因體關聯性分析,並於independent cohort 進行驗證。 研究結果:第一部分橫斷性分析,排除資料不全、目前有抽菸習慣並戒菸不滿一年、極端值的個案後,納入進行常模建立有男性8764人和女性19905人,年齡中位數值為男性51.7歲、女性50.8歲,使用Quantile regression所建立的常模在預測極端值的驗證以及交叉內部驗證皆得到較高的準確度,因此常模選用以Quantile regression所建立的公式。慢性肺疾病以及有抽菸習慣的人,其肺功能每年下降的幅度比預期下降顯著的多,但統計上並無差異。全基因體關聯性研究中發現四個SNPs, rs35517282, rs11122803, rs12376178, rs971889以及CFAP77和KCNAB1基因與肺功能的下降相關,CFAP77基因表現於肺部支氣管上皮之ciliated cell,而則KCNAB1基因參與肺部血管壓力的調控。 結論: 這是第一篇利用台灣代表性的資料以及嚴謹的流行病學方法,建立屬於台灣人肺功能常模的研究,我們發現Quantile regression是最合適於常模的建立的方法。在肺功能下降的全基因體關聯性分析中發現四個SNPs以及CFAP77、 KCNAB1基因與成人肺功能下降相關。 Background Pulmonary function varies by ages. Both genetic and environmental factors influenced the decline of pulmonary function. Researchers interpreted the spirometric indices by percentage of predicted value (% predicted) and the reference value differs from age, sex, height, and ethnic origin. We aim to generate the reference equation of pulmonary function indices and elucidate the SNPs contributing to pulmonary function decline in Taiwanese population. Materials and methods Taiwan Biobank is a large-scale population-based representative cohort in Taiwan. Participants were aged 30 to 70-year-old without cancer diagnosis. Participants with first pulmonary function results and smoking status were enrolled as baseline population; those with follow-up pulmonary function results were left for longitudinal analysis. Generalized additive models for location, scale and shape (GAMLSS) and quantile regression were used to generate predictive equations for FVC and FEV1. Model validation processes included extremely phenotype prediction and ten-fold internal cross-validation. Longitudinal analysis was for investigating environmental risk factors associated with pulmonary function decline. Genome-wide association study with additive genetic model were performed for SNP investigation. The found SNPs were replicated in an independent cohort. Results After removing smokers and unreliable data, there were 8764 men and 19905 women enrolled at baseline. The equations generated from GAMLSS and quantile regression models showed better fitting, compared to Pan 1997 and GLI 2012 equations. After validation, we chose formulae for pulmonary function reference generated by quantile regression model. We found chronic obstructive pulmonary disease would significantly decrease pulmonary function annually than predicted. Smoking did not significantly accelerate pulmonary function decline. Besides, three novel SNPs, rs35517282, rs11122803 and rs12376178, were validated to be associated with accelerating FVC decline. SNP rs12376178 was located in CFAP77 gene region and CFAP77 expressed selectively in ciliated cell of human bronchial epithelium. SNP, rs971889 was found associated with FEV1 annual decline. SNP, rs971889 located in KCNAB1gene region and plays a role in pulmonary vascular tone regulation. Conclusions This is the first study using Taiwanese representative dataset and comprehensive methodology to generate pulmonary function predictive equation. Quantile regression model was the best suitable method. GWAS for pulmonary function decline revealed four novel SNPs, rs35517282, rs11122803, rs12376178, and rs971889, and two novel genes, CFAP77 and KCNAB1 contributing pulmonary function decline. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/68937 |
DOI: | 10.6342/NTU201703655 |
全文授權: | 有償授權 |
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 | |
---|---|---|---|
ntu-105-1.pdf 目前未授權公開取用 | 1.58 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。