Statistical challenges for genome-wide association studies of suicidality using family data

The etiology of suicide is complex in nature with both environmental and genetic causes that are extremely diverse. This extensive heterogeneity weakens the relationship between genotype and phenotype and as a result, we face many challenges when studying the genetic etiology of suicide. We are now in the midst of a genetics revolution, where genotyping costs are decreasing and genotyping speed is increasing at a fast rate, allowing genetic association studies to genotype thousands to millions of SNPs that cover the entire human genome. As such, genome-wide association studies (GWAS) are now the norm. In this article we address several statistical challenges that occur when studying the genetic etiology of suicidality in the age of the genetics revolution. These challenges include: (1) the large number of statistical tests; (2) complex phenotypes that are difficult to quantify; and (3) modest genetic effect sizes. We address these statistical issues in the context of family-based study designs. Specifically, we discuss several statistical extensions of family-based association tests (FBATs) that work to alleviate these challenges. As our intention is to describe how statistical methodology may work to identify disease variants for suicidality, we avoid the mathematical details of the methodologies presented.