Abstract Acute respiratory distress syndrome (ARDS) is a common and fatal clinical condition characterized by inflammatory lung injury, airspace edema, and hypoxemia. The syndrome occurs in an estimated 190,000 patients annually and complicates approximately 24% of admissions to the intensive care unit for mechanical ventilation. The estimated mortality of ARDS is between 35-40%, yet there are currently no effective pharmacological interventions targeted at the disease pathophysiology. Development of disease-specific therapy has been limited by the heterogeneity of ARDS. The identification of biologically-defined subgroups in ARDS may facilitate a precision medicine approach for novel ARDS therapy. The receptor for advanced glycation endproducts (RAGE) pathway is a promising target for ARDS therapies. RAGE is a multiligand pattern recognition receptor on alveolar epithelial cells that amplifies immune and inflammatory responses. RAGE ligands such as HMGB1, AGE, S100A12, and other damage-associated molecular patterns provoke inflammatory signal propagation via RAGE. The soluble form of RAGE (sRAGE) has been implicated in impaired alveolar fluid clearance and is a leading biomarker for ARDS risk and mortality in human studies. RAGE modulating therapies have been shown to improve acute lung injury in animal models, but the translation to human ARDS is challenging. We applied Mendelian randomization (MR), a genetic causal inference method, to plasma sRAGE to assess the association with ARDS risk. Our MR method identified a potential causal relationship between plasma sRAGE and ARDS risk. However, it is unknown whether plasma sRAGE directly, or indirectly via other pathways that shape plasma sRAGE levels, drive the observed effect. Our global hypothesis is that the RAGE pathway causally confers risk for sepsis-associated ARDS, and a RAGE molecular signature can define a sepsis subgroup at high risk for ARDS. This application will use a cohort of septic subjects to characterize the causal contributions of RAGE-cleaving matrix metalloproteinases (MMPs) and RAGE ligands to ARDS risk, and develop predictive utility thresholds to enrich future clinical trials with patients likely to benefit from anti-RAGE axis therapies. The long-term goal of this research is to develop a precision medicine approach to ARDS therapy based on the RAGE axis. Aim 1 examines the role of RAGE-cleaving MMPs in conferring risk for ARDS using MR and mediation analysis. Aim 2 evaluates the causal contributions of the major RAGE ligands to ARDS risk using MR. In Aim 3, we will develop and validate a predictive model that employs RAGE-axis- based thresholds to identify a high-risk population for ARDS in two independent cohorts of critically ill patients with sepsis. The predictive model from Aim 3 may be used for population enrichment in future clinical trials of anti-RAGE axis therapies. The specific aims, training objectives, and structured mentorship in this proposal will bestow th...