From Literature to Hypotheses: An AI Co-Scientist System for Biomarker-Guided Drug Combination Hypothesis Generation
CoDHy is an AI system that connects biomarkers to drug combinations by building a knowledge graph from structured and literature data. It uses graph reasoning and multi-agent systems to generate, validate, and rank drug hypotheses. The system proves that integrating structural knowledge with embedding similarity improves hypothesis quality over LLM-only methods.