Structures of ligand protein complex obtained from X-ray crystallography or cryo-EM often lack the information regarding intermediate stages of the binding-unbinding pathway between solution phase to the final binding pose. Understanding the binding-unbinding pathway is key to understanding ligand selectivity for specific subtypes, a feature needed for the next generation of GPCR drugs. Molecular dynamics simulations could bridge this gap. However, current conventional molecular dynamics simulations have very short timescales compared to binding and unbinding events. The timescale gap made it difficult to observe the dynamic pathway between solution phase and bound ligand, therefore the application of such methods are computationally very expensive. Other methods that use biased energy such as metadynamics often unrealistic conformations that prevents the trajectories to be used as SBDD visual aid.

We employed parallel cascade selection molecular dynamics to elucidate the binding pathway of neurotensin 1 receptor peptide ligands without the use of biased energy methods. We demonstrated the replication of bound ligand conformation in published structures of the neurotensin 1 receptor from solution phase ligand. The Markov model derived from simulations identified key intermediary conformations of the binding pathway that could be translated to drug design strategies.