Robertson, M. J. et al. Structure 28, 707–716.e3 (2020).

Recent developments in cryoelectron microscopy (cryo-EM) now routinely allow near-atomic and atomic-level resolution for biomolecules. Determining structures of protein–ligand complexes, however, remains a challenge, with the resolution of the bound ligand being significantly lower than that of the protein. Thus, understanding ligand binding position necessitates creative solutions, such as using computational chemistry methods in conjunction with experimental data. Robertson et al. have developed GemSpot, an automated computational docking pipeline for modeling and evaluating ligand binding poses in cryo-EM maps. It uses the tool GlideEM to for docking, taking into account the EM map potentials as restraints. Model refinement is done using the OPLS3e force field and quantum mechanics calculations, and water molecule sites are predicted using JAWS. The authors demonstrate the pipeline on a several proteins at varying levels of modeling complexity and EM map resolution.