A multi-agent system for automating scientific discovery
Nature • • business
Key Points:
- Robin is the first multi-agent AI system that fully automates hypothesis generation and data analysis in experimental biology by integrating literature search and data analysis agents.
- The system demonstrated its capability by identifying and validating therapeutic candidates for dry age-related macular degeneration (dAMD), a leading cause of blindness.
- Robin proposed enhancing retinal pigment epithelium phagocytosis as a treatment strategy and identified ripasudil, a ROCK inhibitor not previously linked to dAMD therapy, confirming its efficacy in vitro.
- Further RNA-seq experiments guided by Robin revealed upregulation of ABCA1, a lipid efflux pump, suggesting a novel therapeutic target.
- This semi-autonomous AI-driven approach represents a new paradigm in scientific discovery, enabling iterative hypothesis generation, experimentation, and data analysis within a lab-in-the-loop framework.