AI Science Assistant Robin Autonomously Discovers New Drug Candidates for Eye Disease

Scientists have developed the first artificial intelligence system capable of fully automating the scientific discovery process, from generating hypotheses to analyzing experimental results. The AI system, called Robin, successfully identified and validated novel therapeutic candidates for dry age-related macular degeneration, the leading cause of blindness in the developed world.

Robin represents a breakthrough in AI-driven scientific research by integrating literature search agents with data analysis capabilities. The system can generate hypotheses, propose experiments, interpret experimental results, and generate updated hypotheses in a continuous cycle of discovery. All hypotheses, experimental directions, data analyses, and figures in the research report were produced autonomously by Robin.

By applying this system to medical research, Robin proposed enhancing retinal pigment epithelium phagocytosis as a therapeutic strategy for dry age-related macular degeneration. The AI identified and confirmed the effectiveness of ripasudil, a clinically-used Rho kinase inhibitor that has never previously been proposed for treating this form of blindness.

To understand how ripasudil works, Robin then designed and analyzed a follow-up RNA sequencing experiment. The results revealed upregulation of ABCA1, a lipid efflux pump that represents a possible novel therapeutic target. This demonstrates Robin's ability to not only identify promising treatments but also investigate the biological mechanisms behind their effectiveness.

The research, published in Nature, establishes a new paradigm for AI-driven scientific discovery. As the first AI system to autonomously discover and validate novel therapeutic candidates within an iterative laboratory framework, Robin could accelerate the pace of medical research and drug development by automating traditionally time-intensive processes that require human expertise.