Project Summary Progressive fibrosing interstitial lung disease (PF-ILD) is a devastating condition characterized by relentless lung function decline and survival worse than most cancers. PF-ILD can complicate any ILD subtype and leads to similarly poor survival irrespective of ILD etiology. The inability to predict PF-ILD drives treatment disparities and remains a critical gap in knowledge in those with non-idiopathic pulmonary fibrosis (IPF) forms of ILD. To improve non-IPF ILD outcomes, tools to predict PF-ILD are urgently needed. By applying machine learning algorithms to proteomic data, our group recently developed a preliminary proteomic signature to predict PF-ILD. While promising, notable deficiencies prevented clinical implementation of this signature, including insufficient test performance to justify clinical adoption and the imprecise nature of semi-quantitative data, which limits external generalizability. This proposal aims to address these deficiencies, expand our understanding of PF-ILD and conduct a comparative analysis of validated PF-ILD biomarkers. In aim 1, we will identify and validate semi- quantitative proteomic biomarkers of PF-ILD, hypothesizing that unbiased proteomic investigation will identify novel biomarkers of PF-ILD outcomes and new therapeutic targets. Biomarkers associated with one-year ILD progression and three-year transplant free survival will be identified separately and validated in an independent cohort from the Pulmonary Fibrosis Foundation. Pathway and network analysis will be performed to identify key biologic mediators of PF-ILD and identify new potential therapeutic targets. In aim 2, we will develop a high- fidelity proteomic signature of PF-ILD using a custom, quantitative proteomic platform, which we hypothesize will reliably predict one-year ILD progression when applied prospectively. Test performance characteristics for this signature will be assessed in three prospectively recruited non-IPF ILD cohorts from the US and Canada. In aim 3, we will compare test performance between proteomic, genomic and radiologic biomarkers of PF-ILD, hypothesizing that protein biomarkers will better predict near- and long-term ILD outcomes than genomic and radiologic biomarkers in a prospectively recruited, multi-center non-IPF ILD cohort. Successful completion of this proposal will identify novel biomarkers of PF-ILD that may one day serve as therapeutic targets and lay the foundation for clinical implementation of a proteomic signature to predict PF-ILD. This proposal has high potential to positively impact non-IPF ILD outcomes by allowing patients with PF-ILD to be identified earlier in the disease course and receive prompt intervention to prevent progression.