Dr. Pierre Larmande is a Research Director (DR) at the Institut de Recherche pour le Développement (IRD). He is based in Montpellier, France, in the CERES team (Cereal Root Systems) of DIADE unit. He is also an associated researcher at LIRMM in the ADVANCE team. His research lies at the crossroads of bioinformatics, knowledge graphs and artificial intelligence. His goal is to decipher gene function and regulation in cereals such as rice, sorghum and millets.
His work follows three complementary axes. First, he combines knowledge graphs and AI (graph embeddings, link prediction) to prioritize candidate genes for agronomic traits. Second, he benchmarks genomic language models to predict cis-regulatory elements and gene regulatory networks in plant genomes. Third, he develops graph neural networks for protein function prediction, integrating sequences, interaction networks and ontologies. He also develops NLP and named-entity recognition methods to extract knowledge from the scientific literature.
He is the scientific lead of the AgroLD Knowledge Graph, which integrates heterogeneous genomic and agronomic data for the plant sciences. He coordinates the ANR DIG-AI project (2023–2026) on genotype–phenotype interactions using knowledge graphs and AI. He also contributes to the PEPR Agroécologie et Numérique projects BreIF and AgroDIV, and to the ANR SorDrought project on sorghum drought tolerance.
Dr. Larmande holds a PhD in Computer Science (University of Montpellier, 2007) and a Habilitation (HDR, 2020). He previously led the data integration work package of the Institute of Computational Biology (IBC, 2013–2017). He then co-directed the ICT Lab at USTH in Hanoi, Vietnam (2017–2020). He has supervised over 30 master’s students, as well as PhD students and postdocs, mostly at the interface of computer science and biology. He is an editor for BMC Genomic Data. He regularly serves on program committees (ECCB, EKAW, SWAT4HCLS, INFORSID) and as a reviewer for journals such as NAR, Bioinformatics and GigaScience.
His recent work includes graph neural networks for plant protein function prediction, the evaluation of genomic foundation models for cis-regulatory element prediction, NER models for biology (GRU-SCANET, Bioinformatics Advances, 2025), and representation learning for entity alignment in knowledge graphs (Semantic Web, 2025; ACM Web Conference, 2025).
-- very short bio --
Dr. Pierre Larmande is a Research Director at the Institut de Recherche pour le Développement (IRD), at DIADE unit, Montpellier, France. He works at the interface of bioinformatics, knowledge graphs and AI. His aim is to understand gene function and regulation in cereals such as rice and sorghum. His research combines knowledge graphs and AI agents for candidate gene discovery, genomic language models for cis-regulatory element prediction, and graph neural networks for protein function prediction. He is the scientific lead of the AgroLD Knowledge Graph. He coordinates the ANR DIG-AI project. He holds a PhD in Computer Science (2007) and an HDR (2020) from the University of Montpellier.