Speaker
Description
Over the past four years within the INTER-EXCELLENCE II grant project LUAUS23262, we have developed an integrated ecosystem of open-source tools for computer-aided drug design, covering virtual screening, molecular simulation, de novo design, and ligand optimization. EasyDock provides automated and scalable molecular docking workflows, while StreaMD supports standardized molecular dynamics and free-energy calculations. In parallel, we substantially extended the CReM framework for chemically reasonable molecular transformations and developed specialized tools for pharmacophore-guided generation (CReM-pharm), docking-guided de novo design (CReM-dock), and docking-driven ligand optimization (CReM-opt).
A central objective of this work was to create modular, interoperable, and computationally efficient tools that can be used independently or combined into higher-level automated workflows. Their practical utility was evaluated in two prospective drug-discovery studies. For CDK16, de novo generated compounds were used to guide similarity searching in an ultra-large chemical space of Enamine REAL DB (~10B compounds), followed by experimental testing of selected candidates resulted ion discovery of 14 new CDK16 inhibitors. A second prospective study targeting DYRK1B combined pharmacophore-guided generation, docking-guided design, iterative optimization, physicochemical and structural filtering, and free-energy calculations to prioritize novel compounds for synthesis. Top selected compounds are currently undergo experimental evaluation.
These studies illustrate how open-source computational tools can be integrated into reproducible end-to-end workflows that connect large-scale virtual screening with molecular generation and prospective experimental validation.