声明
摘要
Abstract
Denotation
Chapter 1 Causal Diagrams Theory
1.1 Causal diagrams
1.2 Path implied in causal diagrams
1.3 The d-Separation Rules Linking Causal Assumptions to Statistical Independencies
1.4 do-calculus proposed by Judea Pearl
1.5 Back-door criterion and Front door criterion
1.6 Instrumental variable
1.7 Markov Blanket and its algorithms
1.8 Outline of the dissertation
Chapter 2 Theory and Methodology of Causal inference based on Matching and Regression strategy
2.1 Background
2.2 Methods
2.2.1 Abrief introduction to matched case-control designs under causal diagrams
2.2.2 Simulation
2.3 Results
2.4 Discussion
Chapter 3 Omic biomarkers screening strategy based on conditional independence criterion
3.1 Background
3.2 Methods
3.2.1 Markov Blanket-based repeated-fishing strategy(MBRFS)
3.2.2 Simulation
3.2.3 Application
3.3 Results
3.3.1 Simulation results
3.3.2 Application results
3.4 Discussion
Chapter 4 Identification and calculation of pathogenic pathway effect based on do-calculus
4.1 Background
4.2 Methods
4.2.1 Complex network simplification rules(Figure 4-1)and calculation of causal effect
4.2.2 Segmented series multiplication statistic
4.2.3 Non-parametric permutation and bootstrap test
4.2.5 Simulation
4.2.5 Application
4.3 Results
4.3.1 Simulation
4.3.2 Application result
4.4 Discussion
Chapter 5 Conclusions
5.1 Innovations
5.2 Limitations
Appendix
参考文献
Acknowledge
攻读学位期间发表的学术论文
学位论文评阅及答辩情况表
附英文发表论文