About GrapeLLM
An AI research platform for grapevine science
Research positioning
GrapeLLM is a domain-specific research platform for grapevine (Vitis) science and breeding. It is not a general-purpose chatbot: every component is designed around the questions grapevine researchers and breeders actually ask — from gene function and disease resistance to molecular breeding strategy.
Platform capabilities
The platform connects a domain language model with structured gene knowledge and a literature evidence system. The assistant, ZhiPu, sits at the centre and draws on both structured data and retrieved scientific literature to produce answers that can be traced back to their sources.
Applications in grapevine science
Current focus areas include functional interpretation of grapevine genes, disease resistance gene discovery, gene expression analysis across developmental stages, and molecular breeding decision support. The platform is intended for research use by grapevine scientists and breeding programmes.
Models and evidence system
Two conceptual layers underpin the platform. The first is a domain-adapted language model fine-tuned on grapevine scientific literature. The second is a retrieval and evidence layer that links model output back to specific publications, DOIs and passages, so that claims can be verified rather than taken on trust.
Future development
Planned development includes full PN40024 gene annotation integration, literature-grounded evidence retrieval with citation tracing, expanded breeding topic coverage, and deployment of the domain model inference service. Performance metrics will only be published once they have been independently validated.