Explainable AI for interpretation of Small Angle Scattering
The PhD will be conducted in two laboratories at Paris-Saclay: one group with expertise in artificial intelligence developed over many years, MIA-PS (INRAE), and another in the physics of matter – soft matter, biology – MMB-LLB (CEA/CNRS).
Small-Angle Scattering techniques (X-rays, neutrons, light) involve a constantly growing community, particularly active in France, especially in soft matter and biology. The transition of data from reciprocal space to real space is achieved via different models – in which the MMB group is an expert – whether concerning shape – sphere, rod, platelet, polymer chain – or interactions – attraction, aggregation, repulsion, arrangement. Furthermore, more complex structures, such as proteins or irregular aggregates, require computational or empirical approaches. In all cases, the results are not unequivocal. This is particularly challenging for research groups new to the technique.
In this thesis, thanks to MIA-PS's expertise in AI (machine learning, optimization, visualization), the focus will be on developing explainable AI methods. Part of the modeling involves explained mathematical and physical models, while another part relies on so-called "black box" models, which will be progressively explained. The doctoral candidate will have access to data from three use cases provided by the LLB, and to their experts, to develop a generic methodology. A first step could be based on the globally shared software SasView, a treasure trove of explicit models. We have already received a positive response from the SasView developers, which could therefore serve as a dissemination tool. A valuable contribution will be the access to complementary DPA measurements via the LLB platforms and the SOLEIL and ESRF synchrotrons.
Subsequently, a component focusing on human-computer interaction—ensuring that users remain fully responsible for constructing a physico-chemical-biological explanation—can be implemented. MIA-PS is also an expert in advanced interactive visualization methods.
This project therefore combines highly advanced developments in computer science with a wealth of real-world systems chosen for their originality and, of course, their potential applications.
Endothelial-fibroblast interactions in diabetic foot ulcer: deciphering the intercellular communication responsible for the chronic wound persistence
Diabetic foot ulcer (DFU), a severe complication of diabetes affecting approximately 18.6 million people worldwide each year, is associated with high rates of amputation and mortality. Like other chronic wounds, DFUs exhibit impaired healing due to a dysregulated cascade of cellular signalling and behavioural events that normally ensure rapid closure of the skin barrier. Among the key cellular players, fibroblasts and endothelial cells are central to the proliferative and remodelling phases of wound repair – processes that are notably dysfunctional in chronic wounds. Although endothelial-fibroblast crosstalk is recognized as an essential driver of normal skin healing, the specific mechanisms governing their interaction in DFU is poorly understood.
The main objective of this PhD project is to decipher the intercellular communication between endothelial cells and fibroblasts that underlies the chronicity of DFU. Particular attention will be devoted to extracellular vesicle-associated microRNAs (miRNAs), which are pivotal regulators of intercellular communication through modulation of gene expression in recipient cells. By characterizing the repertoire of pro- and anti-healing miRNAs exchanged between endothelial cells and fibroblasts, this project seeks to uncover novel molecular targets and therapeutic strategies to promote wound repair in diabetic foot ulcers.
V-SYNTHES-guided discovery of BET bromodomain inhibitors : a novel antifungal strategy against Candida auris
New antifungal strategies are urgently needed to combat Candida auris, an emerging multidrug-resistant fungal "superbug" responsible for severe hospital outbreaks and high-mortality infections. Our previous proof-of-concept studies in Candida albicans and Candida glabrata established that fungal BET bromodomains – chromatin-binding modules that recognize acetylated histones – represent promising new antifungal targets. We have developed an advanced set of molecular and cellular tools to accelerate antifungal BET inhibitor discovery, including FRET-based assays for compound screening, humanized Candida strains for on-target validation, and NanoBiT assays to monitor BET bromodomain inhibition directly in fungal cells.
This PhD project represents the translational next phase of our research program. It will exploit the AI-guided V-SYNTHES drug discovery approach – a cutting-edge virtual screening and design framework – to develop highly potent BET inhibitors targeting C. auris. Inhibitors will be profiled in biophysical, biochemical and cellular assays, structurally characterized in complex with their bromodomain targets, and validated for on-target activity in C. auris and antifungal efficacy in animal infection models. They will also be used to explore the emergence of resistance to BET inhibition. This project combines an original antifungal strategy with an innovative methodological approach, offering a unique framework for training in interdisciplinary and translational research.
The combined effects of hypoxia and matrix stiffness on the pathophysiology of pulmonary fibrosis.
Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, and fatal lung disease characterized by excessive extracellular matrix (ECM) deposition, increased tissue stiffness, and localized hypoxia. These alterations disrupt cell–cell interactions within the alveolo-capillary barrier and drive fibrotic progression. This project aims to investigate, under controlled in vitro conditions, the combined impact of mechanical stiffness and hypoxic stress on the fate and phenotype of pulmonary cell types and their intercellular communication. To achieve this, biomimetic polyacrylamide hydrogels with tunable stiffness and specific ECM protein coatings will be developed to support the co-culture of alveolar epithelial cells, endothelial cells, fibroblasts, and macrophages. Cellular responses will be assessed through proteomics, imaging, and secretome profiling. The goal is to identify key mechano- and chemo-dependent pro-fibrotic factors, providing new insights into IPF pathogenesis and opening avenues for targeted therapeutic strategies and lung tissue regeneration.
Studying the structural dynamics of vitamin B12 -dependent photoreceptors in view of biotechnological applications
This integrated structural biology project aims at gaining a mechanistic understanding of the recently discovered family of vitamin B12 -dependent photoreceptors. In particular, we aim at visualising protein conformational changes upon photoactivation from the photochemical timescales (femtoseconds) to the photobiological timescales (milliseconds -seconds). To do so, we will use time-resolved X-ray crystallography and X-ray solution scattering at X-ray free electron lasers (XFEL) and at synchrotrons. By establishing the modus operandi of these newly discovered B12 photoreceptors we will open a window to their rational modification for biotechnological exploitation as optogenetic components.
PtSeipin : linking lipid droplets biogenesis and degradation in the diatom Phaeodactylum tricornutum
Microalgae encompass a wide diversity of organisms and have attracted increasing interest due to their ability to produce biomolecules of biotechnological and industrial relevance. In particular, they can accumulate oil within lipid droplets (LDs) in response to abiotic stresses such as nitrogen deprivation. This oil accumulation holds great potential for biofuel production.We recently demonstrated that knockout of the gene encoding Seipin, a protein involved in LD biogenesis, leads to a strong oil accumulation in the diatom Phaeodactylum tricornutum. Moreover, this accumulation appears to result from an absence of LD degradation in the Seipin-deficient mutants. These findings suggest that, in this diatom, LDs are programmed to undergo degradation soon after their formation, thus inhibiting LD degradation could prove a promising strategy to increase their oil content.This project aims to elucidate the mechanisms underlying LD degradation and, more specifically, the connections between their biogenesis and degradation. Three main research axes will be pursued:
1. Identify PtSeipin interaction partners involved in LD degradation, using both candidate-based and unbiased approaches.
2. Characterize the LD degradation pathways disrupted in PtSeipin knockout mutants by combining electron microscopy with transcriptomic and proteomic analyses.
3. Investigate how microalgae utilize oil during the recovery phase after stress, through fluxomic approaches.
Development of a Modular Enzymatic Platform for the In Silico Design and Synthesis of Novel Therapeutic Peptides via Protein Splicing
The rise of antimicrobial resistance (AMR) has developed into a slow-moving epidemic, fueled by the overuse and misuse of antibiotics, coupled with a stagnation in the development of new antimicrobial agents over the past four decades. Addressing this crisis requires not only more judicious use of existing antibiotics but also the development of innovative drugs capable of overcoming resistant pathogens. In this context, the abundant genomic data generated in the omics era has facilitated the resurgence of natural products as a vital source of novel compounds. Among these, natural peptides—with their unique and diverse chemical properties—have garnered particular interest as potential antibiotics, anticancer agents, and inhibitors targeting specific pathological processes.
The aim of this PhD project is to develop a novel, modular enzymatic tool that enables the in silico design and synthesis of peptides with unprecedented chemical diversity. Central to this approach is the exploitation of a unique chemical reaction: protein splicing. This innovative reaction allows precise removal or editing of specific peptidic sequences, thereby providing a powerful platform to generate hybrid peptides with tailored functionalities, including potential therapeutic agents.
This project will integrate structural and functional studies, computational peptide design and enzyme engineering, aiming to expand the chemical and functional diversity of peptide-based molecules. The successful candidate will work in a state-of-the-art research setting, equipped with cutting-edge facilities and collaborative opportunities, fostering innovative approaches and impactful contributions to the field.
Magneto-mechanical stimulation for the selective destruction of pancreatic cancer cells while sparing healthy cells
A novel approach for selectively destroying cancer cells is being developed through a collaboration between the BIOMICS biology laboratory and the SPINTEC magnetism laboratory, both part of the IRIG Institute. This method employs magnetic particles dispersed among cancer cells, which are set into low-frequency vibration (1–20 Hz) by an applied rotating magnetic field. The resulting mechanical stress induces controlled cell death (apoptosis) in the targeted cells.
The effect has been demonstrated in vitro across various cancer cell types—including glioma, pancreatic, and renal cells—in 2D cultures, as well as in 3D pancreatic cancer spheroids (tumoroids) and healthy pancreatic organoids. These 3D models, which more closely mimic the structure and organization of real biological tissues, facilitate the transition to in vivo studies and reduce reliance on animal models. Preliminary findings indicate that pancreatic cancer cells exhibit a higher affinity for magnetic particles and are more sensitive to mechanical stress than healthy cells, enabling selective destruction of cancer cells while sparing healthy tissue.
The next phase will involve confirming this specificity in mixed spheroids (containing both cancerous and healthy cells), statistically quantifying the results, and elucidating the mechanobiological mechanisms underlying cell death. These promising findings pave the way for an innovative biomedical approach to cancer treatment.
Regulation of gene expression by acetylation and lactylation of histone proteins
In eukaryotic cells, DNA wraps around histone proteins to form chromatin. Dynamic modification of histones by various chemical structures allows for fine regulation of gene expression. Alterations in these complex regulatory mechanisms are responsible for many diseases. Acetylation of histone lysines is known to induce gene expression. Other structures can be added to histones, whose effects on transcription remain largely unclear. Most of them, such as lactylation discovered in 2019, depend on cellular metabolism. We are studying this new modification in murine spermatogenesis: this process of cell differentiation is an ideal model for studying transcription regulation, due to dramatic changes in chromatin composition and gene expression patterns. We have established the distribution of acetylated and lactylated marks on three lysines of histone H3 across the genome. The aim of this thesis is to contribute to deciphering the “histone language,” first by studying the role of lactylations on the transcriptional program. Next, the prediction of chromatin states will be refined by integrating our new data with numerous available epigenomic data within neural network models.
Real-space fitting of flexible molecular structures in high-speed AFM topographic movies
Structural biology seeks to understand the function of macromolecules by determining the precise position of their atoms. Its traditional methods (X-ray crystallography, NMR, electron microscopy), although effective, offer a static view of macromolecules, limiting the study of their dynamics. A new paradigm is emerging: integrative structural biology, combining several techniques to capture, among other things, molecular dynamics. However, despite improvements in femtosecond serial crystallography, molecular dynamics simulations, and cryo-electron tomography, current methods struggle to reach the functional time scale (milliseconds to seconds).
The advent of new scanning probe microscopy, and in particular the recent development of high-speed atomic force microscopy (HS-AFM), allows molecular movements to be observed on the millisecond scale, but lacks the atomic resolution to revolutionize structural biology. The objective of the proposed topic is to further exploit the use of HS-AFM by modeling detailed atomic structures at the heart of the images obtained. The tasks will be both biophysical and computational, involving the improvement of the existing AFM-Assembly tool, which allows direct spatial adjustment of the atomic coordinates of the target molecule under AFM topography. The aim is to apply this protocol to a new type of big data, namely topographical movies obtained by high-speed AFM.
The thesis will be conducted at the Institute of Structural Biology in Grenoble, within the Methods and Electron Microscopy (MEM) group of the Grenoble Interdisciplinary Research Institute (IRIG). It will be carried out in collaboration with the DyNaMo laboratory in Marseille, which specializes in high-speed AFM data acquisition, as part of a joint ANR funding application.
The scientific interest of the project is major for modern integrative structural biology. The great scientific challenge of the coming years in structural biology is the study and analysis of molecular dynamics, in order to move beyond the current paradigm (instantaneous photography) and participate in the emergence of a new paradigm (real-time movie).