DESCRIPTION (provided by applicant): Adaptive decision making relies on the ability to evaluate and integrate potential rewards. Data from primate electrophysiology and human functional magnetic resonance imaging (fMRI) studies implicate key structures including the orbitofrontal cortex (OFC) and striatum in valuation and reward-driven learning. However, how value signals from distinct rewards are integrated into a decision value remains relatively unknown. My research proposal will investigate the neural underpinnings of value computation and construction. In particular, I will conduct several experiments that will lead to a better understanding of these regions' separate contributions and functional interactions. First, I will delineate contributions of the striatum to making valuations involving only positive information, only negative information, or both (Specific Aim 1). Second, I investigate the influence of the striatum on hierarchical value computations, as necessary for complex real-world decisions (Specific Aim 2). Third, I will identify a causal role of the striatum in the manipulation of valuation processes (Specific Aim 3). Despite the large fMRI literature involving the striatum and other reward-processing regions of the brain, these aims have been unaddressed with standard fMRI analyses, which are generally limited in spatial accuracy and ill-suited to make strong conclusions regarding relationships between brain regions. My proposal will employ analytic techniques drawn from the machine- learning and biophysical literatures that are well-equipped to predict individual behavior, identify fine-scale differences in brain computational and cognitive processes, and draw conclusions about causal relationships between brain regions. These analytic techniques, combined with novel experimental designs, will allow me to make a significant contribution to an important area of the neuroscience literature. This proposal focuses on subjective value construction, deficits of which are endemic in clinical disorders. Previous research demonstrates that depressed or drug-deprived individuals differentially process positive and negative material and rewards; the research proposed here will tie those results to value computation and decisions involving risk. Additionally, a better understanding of how individuals make complex choices should prove useful for treatment of mental health disorders like depression or anxiety.