PGCopilot: an AI-assisted research agent for plant functional genomics.

PGCopilot, an AI-assisted research agent for plant functional genomics, integrates planning and tool orchestration to retrieve and synthesize evidence from 1.83 million records and 33 million PubMed-indexed articles.

• Why it matters: Multi-omic data and fragmented plant resources create a bottleneck for deriving insights, while general LLMs often hallucinate or oversimplify plant-specific reasoning.
• What they did: The system uses task planning, identifier disambiguation, and Model Context Protocol tool orchestration to ground responses by integrating evidence across heterogeneous biological databases and STRING interactions.
• The result: In expert-guided benchmarking, it outperformed baseline LLMs on task relevance, evidence integration, hallucination control, and research utility, yielding more verifiable citations and explicit cross-database inconsistencies.

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