We conduct problem-oriented computational research, i.e. we select the most suitable methods to solve scientific questions. We also develop our own computational tools. Recurrent scientific questions are related to ligand binding and ligand design, and to protein sequence-structure-function relationships and molecular evolution. Our expertise covers many aspects of chemoinformatics (databases, QSAR and machine learning, ligand-based approaches) as well as molecular modeling (modeling and docking simulations, virtual screens, ligand optimization). Our ligand discovery projects are conducted in collaboration with a network of experimentalists: pharmacologists and molecular biologists, medicinal chemists, structural biologists. We have a long expertise in modeling membrane proteins (GPCRs, ABC and monoamine transporters, membrane-bound pyrophosphatases). In recent years we have lead projects involving data mining and analysis, in particular machine learning and deep-learning technologies. Our expertise encompasses chemogenomic drug-target networks, genomic data, and structural data on protein-ligand complexes (see here and here).
A selection of projects is presented below.