ParametricversusNon-parametricGeneticAssociation.ppt

Parametric versus Non-parametricGenetic Association AnalysisKristel Van Steen, PhD, ScD(kristel.vansteen@ugent.be)Université de Liege - Institut Montefiore Ghent University – StepGen cvba December 18th , 2007;Genetic Association Studies;Terminology;Terminology;Terminology;Genetic Association Studies;Linkage Disequilibrium;Linkage Disequilibrium;Genetic Association Studies;Genetic Association Studies;Complex Disorders;Markers;Genetic Association;Indirect Associations;Statistical Requirements for a Successful Genome-wide Association Study;Study Designs;Example for Required Sample Sizes;The interpretation of r^2;Power – a Statistical Concept;Online Calculators;Type I and Type II errors;Statistical Analysis depends on Study Design ...;;Assessing Association;Human Genetic Disorders;Mendelian Traits;Complex Traits;Genetic Etiology I;Genetic Etiology II;Genetic Etiology III;Genetic Etiology IV;;Association Analysis;Case-control data structure; ;A Pure Epistatic Inheritance Model;Traditional Method suffers;N = 100;Curse of Dimensionality;Traditional Methods suffer;Type of data;Multi-locus Methods;Limitation of Regression;Limitation of Regression;MDR;Two Measures for Selection of Best n-locus model;MDR Steps;Best Multi-factor Models;Model Selection and Evaluation;MDR Analysis Window;Significance of the Final Model;Measures in Selection of Final model;200 cases and 200 controls; 10 SNPs: 1, 2, 3 , …, 10. Disease etiology due to interaction between SNP 1 and SNP 6.;Simulation II;Mean and standard error of the mean calculated from 50 replicates.;Power of MDR in Presence of Genotyping Error, Missing Data, Phenocopy, and Genetic Heterogeneity ;;;Advantages of MDR;Disadvantages of MDR;Issues to Consider;I: Variable Selection;How many combinations are there?;II: Model Selection;III: Interpretation;Interpretation;MDR: To keep in Mind;References for MDR;Acknowledgements;