A generalizable computational platform for early organ vulnerability modeling and therapeutic discovery using systemic omics and imaging

About This Grant

ABSTRACT This project aims to develop a generalizable computational platform that integrates systemic omics, imaging, and genetic data to model early, anatomically localized tissue vulnerability across major organ systems. While traditional biomarkers are widely used in clinical care, they lack regional specificity and mechanistic interpretability, limiting their ability to forecast subclinical dysfunction or guide targeted interventions. This research will deliver a flexible platform that unites early spatial risk prediction, individualized molecular mechanism inference, and therapeutic discovery to support precision health across the brain, heart, lungs, liver, kidneys, pancreas, muscle, and beyond. The platform will link systemic molecular profiles, proteomics, metabolomics, and transcriptomics where available, to spatially resolved imaging-derived phenotypes (IDPs), such as brain connectivity, cardiac strain, or organ-specific tissue composition. It will model genotype–biomarker–IDP interactions to stratify individual-level susceptibility, leveraging polygenic risk scores and common variants that act through intermediate molecular traits. The framework will support therapeutic discovery through a deep learning–based drug repurposing tool that identifies candidate interventions capable of reversing early molecular dysfunction, prioritizing targets linked to inflammation, metabolism, fibrosis, and mitochondrial stress. To ensure broad applicability, the project will harmonize multimodal data across cohorts and organ systems and disseminate standardized pipelines and prediction tools.