La Rochelle Université is recruiting a PhD candidate on a 3-year fixed-term contract.

Title of the thesis project: Artificial intelligence-driven design and development of antimicrobial materials upon transductive/inductive graph neural network approaches for biomedical applications

Cotuelle: Catholic University of Valencia (UCV), Spain. Biomaterials and Bioengineering Laboratory.

Employer description

Since its creation in 1993, La Rochelle University has been on a path of differentiation.

Thirty years later, as the university landscape recomposes itself, it continues to assert an original proposition, based on a strong identity and bold projects, in a human-scale establishment located in an exceptional setting.

Anchored in a region with highly distinctive coastal features, La Rochelle University has turned this singularity into a veritable signature, in the service of a new model. Its research it addresses
the societal challenges related to Smart Urban Coastal Sustainability (SmUCS).

The new recruit will join the Mathematics Image and Application laboratory (MIA Lab).

An information session on the program will be organized on February 12 from 2 pm to 4 pm to provide you with information on eligibility criteria and the recruitment process. To connect to the meeting on Teams, click here.

Context and research project

The rise of antimicrobial resistance (AMR) and the need for new antimicrobial strategies represent urgent challenges in modern medicine. Traditional antimicrobial agents such as antibiotics are increasingly ineffective due to the rapid emergence of resistant pathogens. In this context, the development of novel antimicrobial materials that can overcome these resistance mechanisms is critical. Artificial intelligence (AI), particularly deep learning (DL) approaches such as graph neural networks (GNNs), offer an innovative approach to accelerate the design and optimization of these materials. GNNs is capable of predicting molecular interactions, allowing for the rapid identification of promising compounds and materials with enhanced antimicrobial properties. This PhD thesis project aims to leverage DL, specifically transductive/inductive graph neural network approaches, to design and optimize antimicrobial materials, making the process faster, more efficient, and more targeted, leading to the development of next-generation materials for biomedical applications to combat microbial infections.

The main objective of this PhD thesis proposal is to develop antimicrobial materials through the integration of DL to predict and optimize the antimicrobial efficacy of new compounds and material composites against viruses, bacteria and fungi. Key objectives include:

1. Designing novel antimicrobial materials by applying AI-based models (specifically DL) to identify the most promising molecular structures for antimicrobial activity.
2. Optimization of material properties, including biocompatibility, stability, and antimicrobial efficacy, using data-driven approaches powered by AI.
3. Testing and validation of the developed materials to assess their effectiveness against viruses, bacteria and fungi in in vitro experimentation under biosafety level 2 conditions.
4. Testing that the developed antimicrobial materials are safe for human beings testing their toxicological aspects in in vitro experimentation under biosafety level 2 conditions.
5. Identification of key factors influencing antimicrobial activity to guide the rational design of future materials.

Scientific Challenges

This research faces several significant scientific challenges:

1. For Dl to effectively predict antimicrobial properties, high-quality, diverse datasets of molecular interactions and material properties are required. Gathering and curating these datasets can be challenging.
2. The rational design of antimicrobial materials requires understanding the complex interactions between material properties, pathogens, and the environment. These interactions are difficult to predict without sophisticated AI tools.
3. While the focus is on antimicrobial efficacy, it is also crucial that the materials are biocompatible and stable for biomedical applications. Balancing these factors while maintaining high antimicrobial activity presents a challenge.
4. To integrate AI and experimental validation to translating AI predictions into real-world applications requires extensive experimental validation to confirm the accuracy of the DL predictions.

Methods to Address Challenges

To address these challenges, the project will utilize a multi-disciplinary approach combining AI, material science, and experimental biology and chemistry:

1. Deep learning: GNNs will be trained on large datasets of antimicrobial compounds and their molecular interactions to predict the efficacy of new material designs. The model will learn to identify key molecular features that contribute to antimicrobial activity.
2. Material synthesis: the materials will be synthesized by combining promising compounds or materials predicted by the DL models. These antimicrobial compounds will be integrated into biopolymers or nanomaterials to create composite materials.
3. Experimental validation: the synthesized materials will undergo a series of antimicrobial in vitro tests, including MIC (Minimum Inhibitory Concentration) assays, disc diffusion tests, and viral inhibition assays, biofilm formation, etc. to evaluate their antimicrobial properties.
4. Optimization and iterative design: based on experimental results, the materials will be refined and re-optimized using further AI predictions. This iterative process will allow for the continuous improvement of material properties. The collaboration with the ProtoQSAR company where the PhD candidate will perform a three months secondment will help to explore interactions between antimicrobial particles and the bioactive compounds within the materials, using molecular docking.

Expected Results

This research is expected to yield several significant outcomes:

1. The creation of novel antimicrobial materials with enhanced antibacterial, antifungal and/or antiviral properties that can be applied in various biomedical fields, such as wound healing, medical devices, and drug delivery systems.
2. The development of a predictive AI framework using Dl that can guide the design of antimicrobial materials, reducing the need for trial-and-error experiments and speeding up the material development process.
3. A deeper understanding of the structure-activity relationships that govern the antimicrobial properties of materials, providing insights into how to optimize materials for specific pathogens.
4. A validated approach for integrating AI-driven predictions with experimental testing, enabling more efficient development of future antimicrobial materials.
5. The successful completion of this project will contribute to addressing the global health challenge of antimicrobial resistance and provide a scalable approach for designing innovative materials with specific biomedical applications. By leveraging AI, this research will pave the way for the development of advanced materials that could significantly impact healthcare and the pharmaceutical industry.