The main goal of this project is to develop inhibitors of L,D-transpeptidases. First, screening of various libraries of compounds that are either already available to the project partners or are commercially available will be performed. Active compounds, either of fragment-size or bigger, obtained from the screening or available from the literature will be used as starting point for further development. They will be used as queries for ligand-based virtual screening of libraries of commercially available compounds. Screenings will be performed using different approaches, e.g. shape similarity using ROCS software (OpenEye), 2D fingerprint similarity using Open Babel and 2D or 3D similarity using the Clique algorithm with LiSiCA software (Insilab). In parallel, structure-based virtual screening will be performed, where more accurate and computationally intensive docking and scoring functions will be used in the Glide software (Schrödinger). Pharmacophoric models will also be created by LigandScout (Inte:Ligand). Virtual libraries will be docked to the active sites of the target enzymes and pharmacophores, taking into account also the covalent binding where applicable, and the compounds with the best sum of scores will be selected for biological evaluation.

Compounds obtained from screenings described above will be re-synthesised and used as starting points for further development and optimization. The goal of this hit-to-lead optimization will be to synthesise new lead compounds with improved properties. Medicinal chemistry efforts will include techniques such as hit evolution (optimization of the original hit by design and synthesis of focused libraries of analogues around a given scaffold with different substitution patterns) to improve potency, selectivity and bioavailability. Based on the co-crystal structures or docking positions, favourable substituents will be introduced into the parent molecule to improve the binding affinity to the corresponding target enzyme.