We aim at a better micromechanical understanding of links between processing, microstructures, and mechanical properties of metallic alloys produced by additive manufacturing. We consider three critical aspects, which are the three-dimensional, heterogeneous, and multiscale morphology of microstructures, the presence of residual stresses due to the printing process, and the heterogeneous plastic deformation of materials when solicited mechanically. To do so, we aim at addressing three scientific questions and challenges: What are the elementary plasticity mechanisms at small scale that impact the mechanical response, how to estimate the residual stresses at the scale of grain aggregates and, can machine learning methods be useful to guide the design of new microstructures with better mechanical performance? We develop and use robust and complementary experimental and modeling tools to answer these questions. We propose, respectively, the development of a spectral discrete dislocation dynamics code numerically able to simulate the deformation of small grain aggregates with precipitates and pores, the measurement of residual stresses in grains with high-resolution 3D-EBSD assisted by thermomechanical modeling, and the use of a micromechanical model fed by X-ray 3D imaging to generate a material database for training and testing machine learning algorithms. We will consider at first copper-chromium industrial alloys as a model system for our study. At longer term, our strategy can be extended to more complex high performance aluminum alloys.

The main objective of the CYCADA project is to develop a straightforward access to important nitrogen heterocycles via cycloaddition reactions. Using new starting materials, the synthetic power of organometallic catalysis will lead to original diastereo- and enantioselective transformations. We will also study diastereodivergent pathways to these important scaffolds using metal/organocatalysis dual catalysis.

Estimation-of-distribution algorithms (EDAs) are general-purpose optimization heuristics that iteratively evolve a probabilistic model of the search domain. Shortly after their inception more than 20~years ago, EDAs employed in practice captured complex (multivariate) dependencies among the problem variables via Bayesian networks, allowing them to solve hard optimization problems and to provide insights into the problem space. In contrast, to this date, all theoretical guarantees for such EDAs assume independence (i.e., univariate dependencies) among all problem variables. With this project, we will reduce this massive gap between theory and practice and prove the first theoretical results for such multivariate EDAs. To this end, we naturally extend the known results for univariate EDAs by incorporating increasing degrees of dependence. This constitutes a milestone for the theory of EDAs and for the theoretical domain of randomized optimization heuristics as a whole.

This project addresses the important problem of the development of accurate atomistic simulation models for computer-aided drug design (CADD). Drug discovery and development are very time and resource consuming processes, which are significantly facilitated by CADD methods. Structure-based CADD methods fundamentally rely on scoring functions to evaluate ligand-target affinities. These simple scoring functions are not accurate enough to reproduce ligand-target interactions in the complex heterogeneous protein-solvent environment. In particular, classical force fields with fixed charges, currently widely used in CADD applications, do not account for induced electronic polarization effects important for ligand:protein binding. Today, there is a clear demand for accurate methods to predict ligand affinities in CADD to speed the discovery of new potential drugs. The primary goal of this project is to develop and implement in CADD a new generation of mechanistic models based on solid physical principles. It is related to the recent advances in next generation atomistic simulation models explicitly treating electronic polarizability. In this project I will extend and implement the polarizable force field based on the classical Drude oscillator model to CADD applications. The main obstacle in using the polarizable force fields in CADD is explained by the treatment of the aqueous solvent, with the current implicit solvent models incompatible with the Drude force field. Moreover, existing CADD methods are challenged by treatment of aqueous solvent, which plays a key role in ligand-target binding. The ATMCADD project will contribute to lifting several technological barriers by designing new solvation models combined with the Drude polarizable force field, and developing cutting-edge methods for solvent treatment in CADD. In particular, in the course of this project I will develop new continuum dielectric medium models for implicit solvation combined with the polarizable force field; develop and implement new models for the vdW dispersive non-polar solvation; and develop new cutting-edge methods for the cavity-formation term, which has been completely neglected in previous studies. The ATMCADD project will be done in collaboration with the main developer of the Drude polarizable force field, making the results of this work highly visible and will create a lasting international collaboration. The second ultimate goal of this project is to apply these advanced models to the important problem of the development of new potent anticancer and antimicrobial drugs. Together with experimentalists, I will propose new inhibitors of the myeloid cell leukemia-1 (Mcl-1) and Bcl-xL oncoproteins, both implicated in many cancer diseases. Specifically, we will design inhibitors disrupting protein-protein interactions of these oncoproteins with the BH3 domains of their pro-apoptotic counterparts to normalize the function of the later proteins. We will also contribute a new approach to tackle the acute problem of bacterial drug resistance by designing new classes of antibacterial drugs. In particular, we will target a recently discovered mismatch-repairing enzyme present in microbial pathogens, but absent in humans. Therefore, this ambitious project combines important pharmaceutical goals with methodological improvements in CADD empowered by the recent advances in polarizable atomistic models. Beyond the methodological improvements that can be expected from the project, we will provide computer methods and tools that will be made available to the drug-design community through their implementation in popular programs for molecular simulations and CADD. Importantly, due to the foundational character of the new methods that we will develop, they can be applied beyond CADD applications, to a variety fundamental and applied problems, such as accurate pKa prediction of protein groups and ligands, protein folding, and de-novo protein design.

Ether compounds are ubiquitous in molecules of interest such as therapeutic agents or agrochemical products for example. Actual classical strategies for their syntheses are mainly based on cross-coupling reactions involving Pd or Cu catalysts in association with aryl halides and phenol derivatives as starting materials. These reagents are either toxic or generate unusable waste. In this project, using nickel as an Earth abundant, non-toxic and cheap metal, we plan to develop two strategies to limit the toxicity of the reagents and wastes. The first one is based on the use of H2O as a unique oxygen atom donor, thus avoiding the use of phenol derivatives. In the second one, ester derivatives will be used as reagents, leading to the expected ethers products thanks to a decarbonylative process, generating CO as only by-product. To succeed, the project will be divided in classical tasks in the field of homogeneous catalysis (ligands and metal complexes design and synthesis, catalysis and mechanistic studies). Based on preliminary results, nickel catalysts with DPPA-type ligands will be developed and used for the targeted reactions. A strong investment will be devoted to mechanistic studies. Indeed, a deep understanding of the mechanism will allow the access to more efficient catalysts and will probably help to envision new synthetic pathways for other cross-coupling reactions. A strong interplay between the LCC and LCM partners who both possess expertise and know-how in the different fields will guaranty the success of the project.