PROJECT SUMMARY Affective lability (AL) is a common psychiatric symptom characterized by rapid mood fluctuations. Clinically significant AL often begins in adolescence, is present in both psychiatric disorders and the general population, and is a major risk factor for suicide. However, the developmental substrates of AL remain only sparsely described. Prior studies have implicated deficits of top-down regulation between the frontoparietal network (FPN) and the amygdala; the FPN is also known to undergo protracted maturation into young adulthood. However, previous efforts have been limited by methodological obstacles related to underlying biological heterogeneity of the FPN. Cortical networks have typically been studied using standardized network atlases, which assume a 1:1 mapping between structural and functional neuroanatomy across individuals. However, recent studies using precision functional mapping techniques have demonstrated reliable individual differences in functional topography, i.e., the spatial distribution of functional networks on the cortex. Notably, the FPN is both critical for affect regulation and also has the most variable functional topography of any cortical network. This proposal will capitalize upon new machine learning tools to map the FPN on a personalized basis to test the over-arching hypothesis that AL is associated with developmental abnormalities of FPN topography and connectivity. In Aim 1, we will continue to acquire a sample of 30 adolescents and young adults (ages 16-21) with affective lability and 20 matched comparators using smartphone-based digital phenotyping and multi- modal imaging. In Aim 2, we will capitalize upon a recently-completed longitudinal study (n=200, 10-25 years old, mean follow up interval = 5 years) that will allow us to understand how longitudinal changes in personalized FPN topography and connectivity associate with AL. Taken together, this project will provide valuable new insights regarding circuit-level developmental deficits associated with AL, and provide the candidate with superb training in computational psychiatry.