This work introduces an automated end-to-end relative free energy (FE) workflow for calculating binding free energies from SMILES strings, using a non-equilibrium switching approach. The workflow's modularity allows for the exchange of components based on license availability, and it investigates the impact of various docking algorithms on free energy accuracy. Results show that both commercial and open-source docking engines can generate poses leading to good correlation with experimental data.

J. Harry Moore, Christian Margreitter, Jon Paul Janet, Ola Engkvist, Bert L. de Groot, Vytautas Gapsys