I want to explain activity of ligands (IC50 already determined) by binding pose,
As per my experience docking is prediction capacity of docking is very poor and i am unable to explain results. Hence, i did MD simulations using two different PDB structures in order to get a conclusive result. However, i am also inclined to use simulated annealing.
So my query is that which is better approach to predict binding pose
1) using stabilized portion of the trajectory in both simulation and clustering the poses of the combined trajectory with pose in biggest cluster being as most probable binding pose
2) using stabilized portion of the trajectory in both simulation and clustering the poses in individual trajectory and comparing the pose in biggest cluster of the both. if both are almost same then it is probable binding pose
3) use simulated annealing to predict binding pose. if yes then do we have to do it multiple times like in MD
Dear Mr. Sharma,
Determine binding pose in molecular dynamics simulations needs to calculating binding free energy and specially using alanine-scanning analysis.then you can compare your md results with predicted pose from docking. For md simulation I suggest amber package rather than gromacs because of its more accurate methods.
Hope it helps
HHA
- Badges
- Science topic