PhD student in AI for plant gene regulation M/F

Unist; Montpellier; Hybrid; EUR 21,000 - 30,000; Full time

Job summary

IRD Occitanie in Montpellier invites applications for a fixed-term doctoral contract within the DIADE unit to study plant genomics using genomic language models and graph learning. The project aims to predict gene expression in plants by linking cis-regulatory elements to expression profiles and evaluating advanced sequence models.

You will work under PI supervision and collaborate with international partners, using GPUs and open science practices.

Qualifications

• Strong command of Python and a deep learning framework (PyTorch).
• Knowledge of machine learning and deep learning (Transformers, representation learning).
• Experience with graph learning (PyTorch Geometric, DGL) and language models (Hugging Face) is a plus.
• Knowledge of genomics and sequencing data handling (NGS formats, bioinformatics tools).
• Scientific English, written and spoken.

Responsibilities

• Build a reference dataset linking cis-regulatory elements to spatio-temporal expression profiles in Arabidopsis and rice.
• Evaluate genomic language models, both general-purpose and plant-specific, for expression prediction. Compare with classical ML approaches.
• Design graph learning models (GNNs, Graph Transformers) that integrate sequence representations into gene regulatory networks.
• Interpret the models to identify key regulatory modules and assess transfer to other cereals.
• Disseminate results: publications and international conferences; write the PhD thesis.

Skills

Python

Deep learning

Transformers

Graph learning

Bioinformatics

Machine learning

PyTorch Geometric

DGL

Hugging Face

NGS data handling

Education

Master's degree or engineering degree in bioinformatics, computer science, or computational biology

Tools

PyTorch

PyTorch Geometric

DGL

Hugging Face

Job description

Organisation/Company IRD Occitanie Research Field Agricultural sciences » Other Researcher Profile Other Profession Positions PhD Positions Application Deadline 4 Dec 2026 - 00:00 (Europe/Paris) Country France Type of Contract Temporary Job Status Full-time Offer Starting Date 1 Mar 2027 Is the job funded through the EU Research Framework Programme? Other EU programme Is the Job related to staff position within a Research Infrastructure? No

Offer Description

Job type: not applicable (doctoral contract)

Research unit (UMR): DIADE (IRD, CIRAD, University of Montpellier)

Location: Montpellier - IRD centre, 911 avenue Agropolis
Open to contract staff: fixed-term doctoral contract, 36 months

About the employer

An international French scientific research institute, IRD helps strengthen the resilience of societies in the face of global upheavals. It operates in more than 50 countries in Africa, Latin America, Asia and the Pacific, as well as in mainland France and the French overseas territories.

Its research responds in concrete ways to priority needs: climate change mitigation and adaptation, the fight against poverty and inequality, biodiversity conservation, access to healthcare, and attention to social dynamics. Research questions are designed with local stakeholders and communities. Teams combine perspectives, disciplines and knowledge through long-term partnerships. Together, they build robust, high-impact solutions.

IRD champions research that benefits as many people as possible. It shares the results of its projects and puts science at the service of action. In this way, it supports the transformation of societies towards fairer and more sustainable social, economic and ecological models.

• The unit you will join

The DIADE research unit (Diversity, Adaptation and Development of tropical plants; IRD - University of Montpellier - CIRAD) studies plant biodiversity, plant adaptation mechanisms and plant interactions with the environment. It explores plant development mechanisms to design new phenotypes needed in a context of climate change. To this end, the unit's teams combine a continuum of approaches, tools and methods: genome analysis, biochemistry, physiology, microscopy, modelling and data integration.

The unit focuses on plants from tropical regions, where current challenges are greatest: exceptional biodiversity, strong population growth, sustained pressure on the environment and rising demand for crop production. The unit is mainly based in Montpellier, with field activities in Africa, Central and South America, and Asia.

• An attractive mission

You will prepare a PhD thesis aiming to predict gene expression in plants. The project couples genomic language models with machine learning on gene regulatory networks (BioLLM-CRE project).

Under the supervision of Pierre Larmande, IRD Research Director and thesis director, and co-supervised by Mikael Lucas, IRD Research Director and thesis co-director, your main activities will be:

• Build a reference dataset linking cis-regulatory elements to spatio-temporal expression profiles in Arabidopsis and rice.
• Evaluate genomic language models, both general-purpose and plant-specific, for expression prediction. Compare them with classical machine learning approaches.
• Design graph learning models (GNNs, Graph Transformers) that integrate these sequence representations into gene regulatory networks.
• Interpret the models to identify key regulatory modules. Assess their transfer to other cereals.
• Disseminate the results: publications, international conferences, writing of the PhD thesis.

Secondary activities:

• Release the benchmark, code and models as open source.

Take part in the scientific life of the unit (seminars, doctoral training).

• Your future team

You will join an interdisciplinary group within UMR DIADE. It brings together bioinformatics and artificial intelligence (genomic foundational models) and root developmental biology. You will have access to high-performance computing resources (GPUs) and an international research environment.

You have developed the following skills:

• Strong command of Python and a deep learning framework (PyTorch).
• Knowledge of machine learning and deep learning (Transformers, representation learning).
• Experience with graph learning (PyTorch Geometric, DGL) and language models (Hugging Face) is a plus.
• Knowledge of genomics and of sequencing data handling (NGS formats, bioinformatics tools).
• Scientific English, written and spoken.

You demonstrate the following personal qualities:

• Scientific curiosity and interest in interdisciplinary work (computer science and plant biology).
• Autonomy and rigour.
• Team spirit and good interpersonal skills.
• Ability to synthesise and to write.
• Perseverance and adaptability.

You hold a level 7 degree (Master's or engineering degree) in one of the following fields: bioinformatics, computer science (AI, machine learning), computational biology.

No field missions. Occasional travel in France and abroad (conferences, collaborations).

Languages ENGLISH Level Good

Additional Information

Joining IRD

Throughout your career, IRD supports you in developing your skills.

The institute provides a range of tools, such as a digital onboarding programme and access to continuing training, promotion and mobility.

Depending on the activities, IRD offers the possibility of teleworking 1 to 3 days per week.

By joining IRD, you will benefit from:

• 32 days of annual leave + 13 RTT days (for a full-time position)
• A staff restaurant or meal vouchers (depending on the site).
• An optional annual membership of the AOS (IRD staff welfare association): holiday, leisure, sports and cultural services
• Supplementary social protection with MGEN, offering high-quality health and provident cover.
• A pleasant working environment close to nature, served by tram line 5 and accessible by bike, with a car-sharing scheme (mobility agreement signed with the Montpellier metropolitan area).

Eligibility criteria

You hold a level 7 degree (Master's or engineering degree) in one of the following fields: bioinformatics, computer science (AI, machine learning), computational biology.