复杂性状遗传风险预测的统计方法研究
发布时间:2021-01-24 05:03
这篇博士论文主要研究基于复杂性状或疾病的,构建风险预测模型的统计遗传学方法。随着复杂疾病研究的不断深入,研究者检测并收集了大量与复杂疾病相关联的遗传变异(如单核苷酸多态性等)。而基于这些探测到的遗传或环境风险因子而构建成的遗传风险预测模型将推进医学和临床的发展。但是迄今为止,用现有方法构建的遗传风险预测模型的精确度都不理想。而与此同时,全基因组关联分析的发展也激发了研究者基于高维数据构建风险预测模型的兴趣。本论文提出了同时适用于基于现有的风险遗传变异或环境因子的,以及基于全基因组高维序列数据的非参数风险预测模型。此外,该风险预测模型可以考虑到基因与基因,基因与环境之间的互作,从而进一步提高了风险预测的精确度。其中“前向ROC”方法主要适用于病例与对照的序列数据的分析需求,而CORC方法则适用于基于家系产生序列数据的遗传风险预测。论文共分四章。第一章引导和概述了复杂疾病研究的发展和现状。特别阐述了用于检测在疾病形成过程中具有重要作用的遗传因子而展开的全基因组关联分析的发展。在此基础上,本论文介绍了疾病遗传风险预测的发展历史,从最初的孟德尔性状到如今的常见复杂疾病,分析了复杂性状与孟德尔性...
【文章来源】:浙江大学浙江省 211工程院校 985工程院校 教育部直属院校
【文章页数】:82 页
【学位级别】:博士
【文章目录】:
致谢
摘要
Abstract
List of Tables
List of Figures
1 INTRODUCTION
1.1 OVERVIEW OF GENOME-WIDE ASSOCIATION STUDIES
1.2 PREDICTIVE GENETIC TESTING
1.3 ROC CURVE AND AUC VALUE
1.3.1 ROC curve
1.3.2 AUC value
1.3.3 Likelihood Ratios and the ROC plot
2 CONSTRUCTION OF PREDICTIVE GENETIC TESTS ON GENOME-WIDECASE-CONTROL DATASETS
2.1 INTRODUCTION
2.2 METHODS
2.2.1 Optimal ROC Curve
2.2.2 The Forward ROC Method
2.2.2.1 Forward selection algorithm
2.2.2.2 Procedure for handling missing data
2.3 RESULTS
2.3.1 Simulation studies
2.3.1.1 Scenario Ⅰ
2.3.1.2 Scenario Ⅱ
2.3.1.3 Scenario Ⅲ
2.3.1.4 Simulations for bagging cross-validation
2.3.1.5 Comparison with Random Forest
2.3.2 Predictive Genetic Tests for Rheumatoid Arthritis
2.3.2.1 Predictive genetic tests based on currently known risk factors
2.3.2.2 A predictive genetic test formed based on whole genome-wide data
2.4 Discussion
3 CONSTRUCTION OF PREDICTIVE GENETIC TESTS ON FAMILIY-BASEDDATASETS
3.1 Introduction
3.2 The CORC Methods
3.3 Simulations Studies
3.3.1 Scenario Ⅰ
3.3.2 ScenarioⅡ
3.4 Data Application for Conduct Disorder
3.5 Discussion
4 SUMMARY AND CONCLUSIONS
PUBLICATIONS
REFERENCES
个人简历
本文编号:2996627
【文章来源】:浙江大学浙江省 211工程院校 985工程院校 教育部直属院校
【文章页数】:82 页
【学位级别】:博士
【文章目录】:
致谢
摘要
Abstract
List of Tables
List of Figures
1 INTRODUCTION
1.1 OVERVIEW OF GENOME-WIDE ASSOCIATION STUDIES
1.2 PREDICTIVE GENETIC TESTING
1.3 ROC CURVE AND AUC VALUE
1.3.1 ROC curve
1.3.2 AUC value
1.3.3 Likelihood Ratios and the ROC plot
2 CONSTRUCTION OF PREDICTIVE GENETIC TESTS ON GENOME-WIDECASE-CONTROL DATASETS
2.1 INTRODUCTION
2.2 METHODS
2.2.1 Optimal ROC Curve
2.2.2 The Forward ROC Method
2.2.2.1 Forward selection algorithm
2.2.2.2 Procedure for handling missing data
2.3 RESULTS
2.3.1 Simulation studies
2.3.1.1 Scenario Ⅰ
2.3.1.2 Scenario Ⅱ
2.3.1.3 Scenario Ⅲ
2.3.1.4 Simulations for bagging cross-validation
2.3.1.5 Comparison with Random Forest
2.3.2 Predictive Genetic Tests for Rheumatoid Arthritis
2.3.2.1 Predictive genetic tests based on currently known risk factors
2.3.2.2 A predictive genetic test formed based on whole genome-wide data
2.4 Discussion
3 CONSTRUCTION OF PREDICTIVE GENETIC TESTS ON FAMILIY-BASEDDATASETS
3.1 Introduction
3.2 The CORC Methods
3.3 Simulations Studies
3.3.1 Scenario Ⅰ
3.3.2 ScenarioⅡ
3.4 Data Application for Conduct Disorder
3.5 Discussion
4 SUMMARY AND CONCLUSIONS
PUBLICATIONS
REFERENCES
个人简历
本文编号:2996627
本文链接:https://www.wllwen.com/xiyixuelunwen/2996627.html