Hi, I have been trying to redock the cocrystallized ligand of the PPARG protein with PDB ID 4yt1. The best conformation (best binding energy) has a high RMSD relative to the original cocrystallized ligand. I was thinking if I can just use the other conformation to validate the search space since the RMSD value relative to the cocrystallized ligand is less than 2? The following are the files that I am working on.
Are all the key protein-ligand interactions reproduced by any of the poses?
Martin Klvana Hi Martin, I have not checked the interactions yet. I just based everything on the RMSD values. If all the key protein-ligand interactions were reproduced, even if the RMSD values are greater than 2, does that indicate that the search space I used in the protocol is valid?
Robertson Rivera
The search space needs no validation, as it can be defined directly based on the crystallographic position of the compound. The relevant question is if you can reproduce the binding mode of the compound, which is defined by its conformation, and its orientation and translation relative to the protein -- assess them separately and try to figure out which of these attributes contributes to the high RMSD and why. Revise the docking protocol accordingly, or try different program.
Martin Klvana Thank you very much! I have read that aside from the RMSD, the important amino acids for the protein-ligand interactions should also be taken into consideration and should be reproduced when we do redocking of the co-crystallized ligand.
Robertson Rivera
Search space alone is very often insufficient for successful docking. Guided docking is better than blind docking. If possible, set additional constraints (pharmacophore features, such as location of hydrogen bond donor or acceptor, location of aromatic ring, etc.) -- if this is not possible with AutoDock, change the docking program.