Planning and Executing a Genome Wide Association Study (GWAS).
Michele M. Sale,Josyf C. Mychaleckyj,Wei-Min Chen
DOI: https://doi.org/10.1007/978-1-60327-378-7_25
2009-01-01
Abstract:In recent years, genome-wide association approaches have proven a powerful and successful strategy to identify genetic contributors to complex traits, including a number of endocrine disorders. Their success has meant that genome wide association studies (GWAS) are fast becoming the default study design for discovery of new genetic variants that influence a clinical trait or phenotype. This chapter focuses on a number of key elements that require consideration for the successful conduct of a GWAS. Although many of the considerations are common to any genetic study, the greater cost, extreme multiple testing, and greater openness to data sharing require specific awareness and planning by investigators. In the section on designing a GWAS, we reflect on ethical considerations, study design, selection of phenotype/s, power considerations, sample tracking and storage issues, and genotyping product selection. During execution, important considerations include DNA quantity and preparation, genotyping methods, quality control checks of genotype data, in silico genotyping (imputation), tests of association, and replication of association signals. Although the field of human genetics is rapidly evolving, recent experiences can help guide an investigator in making practical and methodological choices that will eventually determine the overall quality of GWAS results. Given the investment to recruit patient populations or cohorts that are powered for a GWAS, and the still substantial costs associated with genotyping, it is helpful to be aware of these aspects to maximize the likelihood of success, especially where there is an opportunity for implementing them prospectively.