A question of standards: What makes a good genetic association study?

Genetic association studies are central to efforts to identify and characterise genomic variants underlying susceptibility to multi-factorial disease. However, a major source of frustration has been the difficulty of obtaining robust replication of initial association findings. Much of this inconsistency can be attributed to inadequacies in study design, implementation and interpretation, with the use of inadequately powered sample groups a paramount concern. Several additional factors have an impact on the quality of any given association study - an appropriate sample recruitment strategy, logical variant selection, minimal genotyping error, relevant data analysis and valid interpretation are all essential contributors to the generation of robust findings. Replication has a vital role in demonstrating that associations observed reflect interesting biological processes rather than methodological quirks. An unbiased view of the evidence for and against any particular association needs to focus on study quality, rather than the headline significance value. This chapter discusses these and related issues predominantly as it relates to candidate genes studies rather than genome-wide association studies.