OMERO embedding and Optimisation of Biom3d, an image analysis tool using deep-learning networks

Neural networks, employing deep learning (DL), have become essential for high-throughput image analysis. However, DL tools should be made accessible to a broader scientific community, including plant biologists.

In line with the FAIR (Findable, Accessible, Interoperable, Reusable) principles, we aim:

• (Obj1) to implement new image analysis functionalities into Biom3d, a modular framework based on DL networks initially designed to quantify nuclear shape and chromatin organization.
• (Obj2) to facilitate the use of Biom3D for end users’, we will improve its interoperability with OMERO, a popular image data management system created by the University of Dundee, with Imaris, a proprietary software from Oxford Scientific, and with the FRACTAL web-platform developed by the bioimaging center in Zürich in order
• (Obj3) to contribute to the design of new set of workflows called ChromTrack3D adapted to microscopists’ needs in image analysis. Finally, we will explore connecting OMERO servers within the AGILE consortium to promote the accessibility and sharing of image datasets in collaboration with the University of Dundee.

You will learn how to implement and train new deep-learning models in the Biom3d framework, develop OMERO interface scripts, run jobs on High Performance Computing (HPC) clusters and develop new functionalities in Biom3d like preprocessing OME-Zarr images.

Université Clermont Auvergne is a middle size University located in a city with a high quality of life, a stimulating scientific environment. Clermont-Ferrand is bordered by the Chaine des Puys, a UNESCO classified volcanic formation with great outdoors activities https://www.uca.fr/en

The PhD position will be part of the Chromatin and 3D Nuclear Dynamics (CODED) team co-lead by Aline Probst and Christophe Tatout at the Institute og Genetics, Reproduction and Development (iGReD), a pluri-disciplinary team working in plant epigenetics and bioimage analysis in a friendly working environment. The team has an international recognition and is leading the network EPIPLANT of the French epigenetic community https://www.igred.fr/en/team/chromatin-and-3d-nuclear-dynamics-coded/ 

Significant experience in Python (and Java) scripting, and knowledge in (plant) molecular biology and image analysis.

A Master in the bioinformatics or image analysis. A desire to succeed in science.

Applicants must

• not have resided OR worked in France during the last three years (or less than 12 months)
• hold a master’s degree (obtained no later than September 2026)