LMDM
The lines below organise what the laboratory investigates. They overlap frequently: a single project usually spans structural modelling, simulation and computational screening.
We develop and adapt tailor-made computational solutions to meet scientific, academic and technological demands.
Many of the laboratory’s questions are answered by no off-the-shelf tool: what is missing is a pipeline that builds homo-oligomeric complexes in batch, a converter between analysis formats, a contact metric that separates contributions per residue. Our tools grow out of that friction and are released as open source, and they are gathered under the Software tab.
We apply computational approaches to identify, analyse and optimise bioactive molecules in the context of rational drug design.
In practice this combines molecular docking, virtual screening of compound libraries, molecular dynamics simulations of the resulting complexes and binding free-energy estimates. The goal is twofold: to prioritise candidates for experimental assay, and to explain, at atomic level, where the observed affinity and selectivity come from.
We explore the molecular mechanisms involved in biological and physicochemical processes through theoretical and computational approaches.
A protein is not a static structure: its function emerges from motion. This line covers protein folding and unfolding, the conformational transitions that open and close binding sites, the behaviour of systems under high hydrostatic pressure, the interaction of proteins with the biomembranes that anchor them, and protein–protein interactions. Because the relevant states are usually separated by barriers that conventional simulation rarely crosses, the work depends on enhanced sampling methods and free-energy calculations.
We use quantum mechanical methods to investigate electronic properties, reactivity and molecular interactions at the atomistic level.
Not every biological phenomenon is described by a classical force field: bond breaking and formation, proton and electron transfer and metal coordination require quantum treatment. Hybrid QM/MM methods make it possible to describe the reactive site at the electronic level while keeping the protein environment and the solvent classical, and also to derive parameters for ligands missing from conventional force fields.
We study the organisation, modelling and structural analysis of proteins at large scale, to understand their functions and interactions.
Most of the laboratory’s questions begin with a structure that does not exist experimentally. We build three-dimensional models by comparative modelling and by machine-learning-based prediction methods, assess their stereochemical and energetic quality, and refine them by simulation before any downstream use. Applied at scale, the same methodology treats a pathogen’s entire proteome as a search space for targets, not only the few proteins already crystallised. Metabolic network analysis complements that search by pointing to the organism’s points of fragility.
