Agent execution layer for biology workflows: batch gene scanning, tumor cohort briefing, target discovery — with structured outputs, resume, and pipeline-ready artifacts
-
Updated
Aug 26, 2026 - TypeScript
Agent execution layer for biology workflows: batch gene scanning, tumor cohort briefing, target discovery — with structured outputs, resume, and pipeline-ready artifacts
Network medicine drug target identification from protein interactome
AI-powered cancer target discovery agent for precision oncology research. Input genomic variants → find matched drugs via CIViC/DGIdb/PubMed → rank by evidence. Built with LangGraph + DeepSeek API + Streamlit.
GNN-based drug-target interaction prediction with binding affinity estimation
[Donor] Open Targets drug target database R client
To associate your repository with the drug-target topic, visit your repo's landing page and select "manage topics."