Pleiotropy in complex traits: challenges and strategies

Abstract

Genome-wide association studies have identified many novel loci for hundreds of traits. Interestingly, numerous genetic loci have been associated with multiple seemingly distinct traits. These cross-phenotype (CP) associations highlight the relevance of pleiotropy in human disease. There is substantial evidence for CP associations in contemporary gene-mapping studies. Different types of pleiotropy (biological, mediated and spurious pleiotropy) can underlie a CP association. Various analytical approaches have been devised for detecting CP associations, especially methods that are based on summary statistics as opposed to individual-level data. Different methods have relative advantages and disadvantages and are distinguished by their underlying algorithms and by the types of phenotype data that they handle. Study design considerations are crucial for minimizing the identification of spurious CP associations. CP associations can highlight shared biological pathways and, when associated with different diseases, have clinical implications for diagnosis, counselling and treatment.