Genetics-First Approaches to Parsing Variability in Neuropsychiatric Disease

Abstract

It has proven hard to identify reproducible and diagnostically-specific biomarkers in psychiatry. Three major contributors to this problem are: the limited biological validity of current diagnostic constructs; the staggering complexity of brain development; and the relative inviolability of human brains. This talk will consider some potentially adaptive responses to these obstacles – centered on the study of naturally-occurring genetic disorders as objectively-defined entry points into the complex biology of neuropsychiatric disease. The work presented will span diverse data types and analytic methods, ranging from deep-phenotypic symptom mapping, to large-scale in vivo neuroimaging, and linkage of neuroimaging data to postmortem maps of the human brain. A recurring theme throughout will be the value of cycling between linked questions in clinical and basic neuroscience. I will first consider how studies of genetically defined neurodevelopmental disorders can guide us towards fundamental basic questions regarding the scaling of human brains and behaviors. We will then see how answers to these basic science questions can help to better pinpoint the selective vulnerability of different brain systems to a range of genetic risk factors. Finally, I will present our ongoing efforts to (i) build multiscale reference maps of the human brain from postmortem data, and (ii) apply these maps to bridge genetic and anatomical scales of analysis in psychiatry. Such multi-level research approaches may help us identify phenotypes that more directly reflect the biology of neuropsychiatric risk in humans – with beneficial consequences for both basic and clinical aspects of biological psychiatry.