Candidate‐gene studies have traditionally selected plausible candidate‐genes based on a theory on the underlying biological mechanisms, for example, relating the dopamine cascade to ADHD 26 or substance use. 27 This approach is limited by current knowledge of the biology of investigated behaviors. 27 In addition, candidate‐gene studies are often restricted by a lack of available data resulting in underpowered or small‐scale designs 28 and examination of only a few (or a single) phenotype(s). 29 Consequently, these limitations have rendered the candidate‐approach largely unsuccessful. 30 , 31
