Multi-Omics Drug Repurposing Platform
Built with Graph Neural Networks on Biomedical Knowledge Graphs
Graph Attention Networks processing multi-relational biomedical knowledge graphs with 5.8M triplets from DRKG.
Predict novel therapeutic uses for existing drugs by analyzing compound-disease-gene interaction networks.
SHAP-ready architecture for interpreting predictions. Understand which molecular features drive each candidate.
Streamlit-powered interface with real-time predictions, network visualization, and downloadable reports.
Combines genomics, proteomics, and metabolomics data through unified graph representations.
End-to-end ML pipeline from raw data ingestion to deployed prediction API with Docker support.