Agentic Discovery Transfer Enables Cross-Cancer Discovery at Scale

SMART, a multi-agent LLM framework for scRNA-seq discovery transfer, tested 121 studies across six cancer types and found a lung cancer prognostic signature linked to survival in liver cancer, pancreatic cancer, and pleural mesothelioma.

• Why it matters: Published cancer findings are seldom validated in other cancer types or patient populations because cross-dataset testing is labor-intensive and hard to scale.
• What they did: SMART takes a source manuscript and a target scRNA-seq dataset, extracts clinical and mechanistic claims, generates and runs analyses for testable findings, then reports which claims transfer.
• The result: It identified cross-cancer transferable findings and produced new mechanistic hypotheses, including confirming the lung signature’s association with survival in liver cancer, pancreatic cancer, and pleural mesothelioma.

AI summary of the abstract · may contain errors · not medical advice · read the paper.