A Whole-Brain Atlas of Genetic Effects on Psychiatric Symptomology

          Anatomical variation in the brain is a candidate endophenotype for psychiatric symptomology. Initial work has linked variation (e.g., cortical thickness, surface area) in anatomical regions of interest (ROIs),... [ view full abstract ]

          Anatomical variation in the brain is a candidate endophenotype for psychiatric symptomology. Initial work has linked variation (e.g., cortical thickness, surface area) in anatomical regions of interest (ROIs), to most psychiatric syndromes, with Insula and dorsolateral prefrontal cortex often predictive of psychiatric outcomes. Although promising initial descriptions, this work does not take advantage of the high dimensionality of fMRI images, i.e. vertex-wise (smallest possible picture of brain area) estimates.  Furthermore, genetic effects may vary within predefined ROIs, and relative differences in heritability between ROI’s may be overestimated. In this study, we map the genetic effects on anatomical measures prediction of symptoms to create whole-brain atlases of genetic effects on psychiatric disorders.   We use a subset of 270 twins from the Colorado Longitudinal Twin Study, with replication from a comparable twin sample from the Human Connectome Project. Initial results of genetic estimates across 148 ROI’s show left hemisphere structures tend to be under larger genetic influence than right structures.  As such, bivariate brain-behavior biometrical models show genes influencing left structures are more predictive of psychiatric variation.  Broadly, psychiatric variation was associated with genes influencing larger somatosensory regions and smaller insula regions. The genetic influences on brain regions that most strongly predict one psychiatric disorder also predicted other psychiatric disorders, suggesting a possible underlying anatomical genotype for psychiatric behaviors.  Prior to presentation, we will map these predictions for multiple psychiatric symptoms at the vertex-wise level.  Our results suggest that anatomical measures are a key endophenotype for the association between genes and disorders.