Dynamic docking in protein-ligand modeling

Large numbers of docking jobs are used in a distributional sense to obtain computational molecular non-covalent and covalent binding information of ligands to protein cavities. A numerically calculated density of binding states is generated and used to identify binding modes, which further can be used to calculate thermal observables. With the density of states the final state non-covalent or covalent protein-ligand entropy can be calculated. Binding strengths, conformations, and individual atomic interactions are found per binding mode as well as the population of binding states that a ligand can take in interaction with a protein cavity. The binding and conformational results are analogous to what is found from molecular dynamics trajectories per protein conformation but uses a GOLD molecular model and docking. The calculation is orders of magnitude more efficient than MD simulations. Brief studies of 2 small molecules, Nirmatrelvir and BAM-15, demonstrate the protocol and the use.