Genetic Association Study Design
Association studies can be performed using any type of genetic polymorphism, but the one most commonly used is the single nucleotide polymorphism (SNP). This is because of its abundance across the genome, comprising 90% of all human variation. The aim of this chapter is to review various aspects of population-based study design, aimed at uncovering common genetic variations underlying complex traits. There is the realization that many of these variants lie ingene deserts and their biological function is not obvious. These variants tend to explain only very small proportions of overall heritability, leaving many wondering where the missing heritability can be found. The first type of genetic association study to be conducted, before the availability of dense SNP LD maps, was the candidate gene study. Most candidate gene studies have suffered from limitations in study design, including inappropriate case definition and control selection and—most importantly—vastly inadequate sample sizes. Despite early concerns about multiple testing issues and power, GWA studies have brought the gene mapping successes for complex traits, which candidate gene studies were largely unable to deliver. Following this, the chapter also presents the population-based study design. The most popular population-based study design for binary traits (whether focusing on candidate genes, regions, or the entire genome) is the case-control study. In a case-control study, a set of cases is identified and their genotyping information is compared to a set of suitable controls in order to find genetic variants associated with the trait under study.