This study will evaluate the feasibility and clinical utility of developing predictive models of treatment response for patients with depressed mood using multimodal clinical data collected in real-world clinical settings. The study will examine outcomes among patients treated with interventions including esketamine, conventional transcranial magnetic stimulation (TMS), or Stanford Accelerated Intelligent Neuromodulation Therapy (SAINT). Retrospective clinical and research data from existing databases may also be incorporated, when available and authorized, to support model development and validation. The goal is to assess whether integrated clinical datasets can be used to support the development of predictive tools that may inform personalized treatment selection in depression.