We propose to test three novel approaches to the molecular genetic bases of psychopathology in children and adolescents:

Specific Aim 1: Functional gene group analysis. It is possible to construct polygenic risk scores that account for useful amounts of variance in phenotypes by summing multiple genetic variants of small individual effect. One method that should be particularly helpful in discovering the neurobiological risk mechanisms of psychopathology is to base polygenic risk scores on the many genetic variants that all influence the same cellular function. For the ALSPAC data, functional gene group scores would be created quite easily by running computer algorithms on the existing ALSPAC SNP data.

Specific Aim 2: Hierarchical phenotypes. We may have failed to find strong genetic associations partly because psychopathology phenotypes have not been measured in ways that accurately reflect the nature of genetic risk. Heretofore, we have mostly studied one categorical mental disorder at a time. We propose to use a novel hierarchy of dimensional mental health phenotypes to optimally identify pleiotropic genetic risk variants.

Specific Aim 3: Environmental moderation and sex differences. Testing for gene-environment interactions (GxE) is not a novel strategy, but it is still controversial, greatly underused, and critically important. Therefore, we will test for GxE with prenatal and postnatal environmental variables assessed prospectively to describe their moderation of genetic associations.