Chronic, clinically-impairing irritability is a serious health problem that affects 3-5% of youth and predicts both adult depressive and anxiety disorders and long-term impairment. Current studies on the neural mechanisms of irritability in youth employ traditional univariate analyses and have limitations: (1) they examine activation patterns in particular regions, thus using only a small portion of the data and (2) they tend to find small to moderate effect sizes. In addition, little is known about functional connectivity at rest and its association with irritability or the ability of neural measures to predict changes in irritability over time. These limitations could be addressed through research using multi-voxel pattern analysis (MVPA), i.e., by applying machine learning methods to both resting state and task-based fMRI data. MVPA exploits the full spatial pattern of brain activity and thus has increased sensitivity compared to conventional univariate, individual-voxel-based general linear model (GLM) approaches. However, univariate analyses may provide complementary or unique information relative to MVPA, given its sensitivity to variability and global engagement of task effects between subjects. The specific aims of this project are to identify patterns of whole-brain activation (Aim 1 & 2a, 2b) and functional connectivity at rest (Aim 2a, 2b) using a combination of MVPA and univariate approaches to predict (1) individual differences in irritability in a clinical sample of youth and in healthy youth with varying degrees of irritability, and (2a) individual differences in irritability and (2b) changes in irritability 6 months later in a community sample of youth with varying degrees of irritability. This work will provide new insights into the pathophysiology of multiple mental disorders for which irritability is an important symptom, and it has the potential to identify biomarkers that can be used to predict the course and prognosis of clinically-impairing irritability. The long-term goal of the PI is to become an independent NIH-funded investigator with a neuroscience-focused research program targeting developmental trajectories of irritability. To achieve this goal, the PI aims to expand her previous training to support both advanced, independent research on resting state and task-based fMRI and the use of pattern classification techniques, as well as their integration with univariate methods, applied in samples with varying degrees of irritability. This training objective directly supports the PI’s research objective, which is to identify the neural mechanisms of irritability and neural correlates that predict changes in irritability over time using a combination of MVPA and univariate imaging methods applied to task-based and resting state fMRI data. The central hypothesis, based on the PI’s pilot data, is that activity patterns during a frustration task, along with connectivity patterns at rest, will predict levels of, and changes in,...