Associating complex traits with genetic variants: polygenic risk scores, pleiotropy and endophenotypes

Genotype–phenotype causal modeling has evolved significantly since Johannsen’s and Wright’s original designs were published. The development of genomewide assays to interrogate and detect possible causal variants associated with complex traits has expanded the scope of genotype–phenotype research considerably. Clusters of causal variants discovered by genomewide assays and associated with complex traits have been used to develop polygenic risk scores to predict clinical diagnoses of multidimensional human disorders. However, genomewide investigations have met with many challenges to their research designs and statistical complexities which have hindered the reliability and validity of their predictions. Findings linked to differences in heritability estimates between causal clusters and complex traits among unrelated individuals remain a research area of some controversy. Causal models developed from case–control studies as opposed to experiments, as well as other issues concerning the genotype–phenotype causal model and the extent to which various forms of pleiotropy and the concept of the endophenotype add to its complexity, will be reviewed.