PROJECT SUMMARY Compulsivity, persisting with set ways of thinking and behaving in the face of negative consequences, is a prominent component of multiple psychiatric disorders. Given the many millions of patients and family members impacted by harms stemming from compulsive behavior, there is an urgent need to better understand the circuit basis of this quintessentially transdiagnostic phenomenon. The objective of this K23 application is to equip the candidate, a psychiatrist with a strong quantitative research background, to carry out intracranial studies of circuit level computational deficits in human subjects with serious psychiatric illness. A new generation of Food and Drug Administration (FDA) approved deep brain stimulation (DBS) devices, precisely targeted and chronically implanted, can both stimulate and record neural activity. Our preliminary data shows that this new technology can be successfully incorporated into the treatment of patients receiving therapeutic DBS for severe obsessive- compulsive disorder (OCD), making it possible to directly record from clinically relevant circuitry as patients go from sick to well. We propose to pair repeat measures of important cognitive processes (inhibitory control, reversal learning) with serial neural recordings of key cortico-basal ganglia-thalamo-cortical (CBGTC) circuit nodes. This experimental design will allow us to ascertain how clinical improvement is accompanied by changes in cognitive processes, and to uncover the relevant circuit activity. In Aim 1, to identify the cognitive changes caused by therapeutic brain stimulation we will repeatedly administer validated assays of flexibility (a reversal learning task) and inhibition (a stop signal task). At each time point we will computationally model relevant cognitive parameters, so that we can track evolution of decision-making dynamics as treatment progresses. In Aim 2, we will identify neural underpinnings of the cognitive processes changed by therapeutic brain stimulation. We have implemented an experimental paradigm that allows for millisecond-precision alignment of behavior data and neural activity. Each time the participants perform a task, local field potentials (LFPs) will be recorded from key CBGTC circuit nodes, never previously accessible for study in patients with compulsive disorders. With the unique opportunity to longitudinally record directly from the basal ganglia in these patients, we will be able to establish if these circuits do indeed subserve clinically relevant decision-making processes. This K23 proposal is supported by a carefully assembled, collaborative, and diverse mentorship team with the requisite expertise in psychiatric DBS, disorders of compulsivity, computational modeling of behavior, and human intracranial recordings. The candidate will emerge from the mentored research experience equipped to lead independent studies of deep brain circuit function in real world patients.