Mots-Clés

image analysis 3D image analysis cryo-electron tomography deep learning python linux

Description

Mission

Develop and implement computational methods for cryo-electron tomography data, applying them to microbial systems to advance our understanding of nanoscale cellular architecture.

Activities

Cryo-electron tomography (cryo-ET) allows visualization of molecular structures within their native cellular environment. However, to fully exploit the advantages of this technique, a major challenge remains: accurately identifying and locating specific protein complexes within the crowded, noisy cellular landscape.

Our research objective is to develop a novel computational framework that represents protein complexes in cryo-ET volumes as graphs, naturally handling the irregular, incomplete, and non-uniform geometry present in biological scenarios. This can be achieved by combining two complementary approaches: density tracing and deep learning methods. The former draws from mathematical methods to produce a skeleton representation of the tomogram that can be naturally interpreted as a graph, while the latter enables the interpretation and analysis of the data within such a framework.

The Ph. D. project will focus on establishing the graph construction pipeline, converting cryo- ET volumes into graph representations for structural detection followed by advanced 3D image analyses adapted to specific biological configurations. Working with synthetic and real data from microbial systems, the candidate will identify, characterize and provide structural insights into macromolecular complexes and higher order assemblies within cells. This foundational work will support future developments in graph neural networks for template-free protein identification and contribute to a deeper understanding of supramolecular organizations in cells.

Skills

This position is designed for individuals with strong interest in image processing and curious about biological systems. Structural and computational biologists, as well as quantitative profiles in mathematics, physics and computer science are welcomed to apply.

-Experience or interest in using the computational tools necessary for image processing (Linux, computing cluster, image processing software).
- Knowledge and experience in Python and deep learning frameworks is highly desirable.
- Interest for biological cryo-electron tomography image analysis and proclivity for learning state-of-the-art cryo-ET image analysis software.
- Problem-solving ability, interest in biology, desire to work in an international and multidisciplinary team.
- Independent thinking, rigor and organization skills.
- English level minimum B2.

Work context

The Institute of Structural Biology (IBS) is a research institute affiliated to the Université Grenoble-Alpes, the CNRS and the CEA. IBS is a major national and international player in the field of integrated structural biology. It currently employs about 300 scientists, including numerous students and postdocs from all over the world, and benefits from state-of-the-art infrastructure and equipment in structural biology, biophysics, biochemistry and cellular and molecular biology. The IBS is one of the founders of the Partnership for Structural Biology (PSB) with prestigious European institutes: the European Synchrotron Radiation Facility (ESRF), the ILL neutron scattering facilities, and the European Molecular Biology Laboratory (EMBL) Grenoble outstation, all located at the same campus, offering a strong international environment. Within the IBS, the candidate will join the “Microscopic Imaging of Complex Assemblies” (MICA) group, interested in high resolution 3D imaging of macromolecular complexes and cells with the goal of gaining structural insights that would yield functional understanding of cellular processes: https://www.ibs.fr/fr/recherche/assemblage-dynamique-et-reactivite/groupe-imagerie-microscopique-d-assemblages-complexes-i-gutsche/?lang=en