PROJECT SUMMARY/ABSTRACT Up to half of depressed older adults do not respond to an initial antidepressant, and fewer than 20% achieve remission. Using a robust multi-site clinical trial network, we have completed two landmark randomized controlled trials (RCTs) showing that augmentation of current antidepressant with aripiprazole (ARI) or bupropion (BUP) results in a 29% remission rate in TRLLD. In response to RFA-MH-24-120, we will use this and other existing clinical trial data to design and evaluate a clinical decision support tool, the Biotype-assigned Augmentation Approach in Resistant Late-Life Depression (BAARD), to improve treatment selection using precision biomedical information. With integrated expertise in geroscience, psychopharmacology, cognition, molecular subtyping, neuroimaging, computational psychiatry, and qualitative research methods represented on our research team, we will use existing demographic, clinical, cognitive, genetic, proteomic, and high quality neuroimaging on ~700 participants to develop and test the BAARD tool. UG3 Phase Specific Aims: Design and Validate the BAARD Decision Support Tool. Aim 1, Yr 1: Develop the BAARD tool for treatment selection in TRLLD using data from two large multi-center trials (‘IRL-GRey’ & ‘OPTIMUM’) and the embedded biomarker study (‘OPT-Neuro’) (total N~700). Aim 2, Yr 2: Refine the BAARD biotype profile & analytic approach using data from individuals diagnosed with major depressive disorder (MDD) or treatment resistant depression (TRD) from the Canadian Biomarker Integration Network in Depression (CANBIND, N~200) and UK Biobank (UKB, N~2,400) studies, respectively. Go-no-go threshold for Yrs 1-2: To proceed to the proposed UH3 RCT phase (N=300), our BAARD tool must achieve a combined predicted remission rate of ≥46% in cross-validation. This corresponds to a balanced accuracy of ≥75% for remission prediction in test data and will be assessed at the end of Yr 1 and Yr 2. UH3 Phase Specific Aims: Test the BAARD Decision Support Tool in a Prospective RCT. Aim 3, Yrs 3-5: In the UH3 phase, we will randomize 300 adults aged ≥60 yrs with TRLLD (2:1) to BAARD tool- assigned treatment vs. randomized treatment assignment (1:1 ARI:BUP) and compare remission rates based on the Montgomery Asberg Rating Scale score (MADRS score<10) following 10 weeks of treatment. Hypothesis: The remission rate in the group treated with ARI or BUP based on BAARD predictions will be at least 17% higher than the remission rate observed in the randomly treated comparison group (e.g., 46% vs. 29%). Go-no-Go threshold for Yrs 3-5: The annual threshold for proceeding with the RCT will be based on participant enrollment milestones (first by end of Q9; 50% by end of Q12; 100% by end of Q15). Exploratory Aim, Yrs 4-5: Obtain stakeholder input on acceptability and appropriateness of BAARD and explore facilitators and barriers to implementation in future clinical trials, and eventually, routine clinical care.