Population genetics, polygenic inheritance, and pharmacogenetics

There has been much to report from genetic epidemiology over the last 12 months. The use of common genetic variants (>1% allele frequency) to understand mechanisms of disease continues to make rapid progress through advances in high-throughput genotyping capacity (see Klein et al. below) and huge study sizes (see Guo et al.). There are also some advances in analytical methodology, such as the concept of Mendelian Randomisation (see Davey-Smith et al.). However, it is still unclear how best to analyse the mountains of genotype data that can now be rapidly generated. Conservative approaches that choose only the most statistically significant results will no doubt throw out many true positive findings. In addition to solely mathematical approaches (to choose the best candidates potentially from all 10 million human single nucleotide polymorphisms (SNPs)), there is increasing debate over statistical methods that include consideration of prior knowledge as to which genes are involved, i.e. a Bayesian approach to analysis (1). A recent perspective on the genetics of type 2 diabetes showed that knowledge of rare mutations that have substantial effects (e.g. which cause severe insulin resistance, maturity-onset diabetes of the young (MODY) or Neonatal Diabetes) can support the findings of studies of common genetic variants (2). The authors concluded that we may not yet close to understanding the complex polygenic aetiology of type 2 diabetes, but at least we may be at the end of the beginning. While several of these studies may be focused on disease outcomes outside of the field of Paediatric Endocrinology, the expected moderate and lifelong effects of common genetic variants on human physiology would predict that many of these genetic effects may be detectable in childhood, and may have much relevance to hormonal activity, growth, and development. From a clinical perspective, the use of DNA analysis for common genetic variants continues to grow, particularly to identify subjects at high risk for drug side-effects (3). There are still no current examples of routine pharmacogenetics in the Paediatric Endocrine Clinic, but if the results of Dos Santos et al. (see below) are replicated, genetic epidemiology may be coming soon to a growth clinic near you!