active · ANR (FR) · SNSF (CH) — ANR-24-CE93-0012-01
MetaboLinkAI
A Franco–Swiss consortium developing an AI agent ecosystem for metabolomics — automating data interpretation, annotation, and knowledge integration across mass-spectrometry workflows.
MetaboLinkAI is a Franco–Swiss consortium co-led by eDIAM / HolobiomicsLab (CNRS · UniCA) with partners at the University of Geneva, ETH Zürich, SIB, and others. It develops a network of AI agents that connect mass-spectrometry data, literature, and structural databases to accelerate metabolite annotation and biological interpretation.
eDIAM contributes the LLM + knowledge-graph backbone, the multi-agent reasoning layer, and the metabolomics-specific evaluation harness. Several team members — Matthieu Ferraud, Martin Legrand, Benjamin Navet — are directly engaged in the consortium's technical development.