[Dock-fans] Thanks Terry and I found out what was wrong......
nicole23 at naver.com
Thu Jul 21 02:42:00 PDT 2005
While inspecting bounded structure of the protein and virtual ligand, I found out that
conformations which gave me stranges scores were almost stretched out. It means that
ligand is in linear form not bent. When I compared it with box pdb file I could understand
what went wrong. The some atoms of ligand was located out of the box! -_-;
I changed the box coordinates and some parameters as Terry advised to me. And it
looks fine until now.
I want to ask you about analyzing results of DOCK. When I execute docking run, I make them run several times with different random seeds to see if the conformation is trapped in some strange local minimum. Most of the time obtained conformations are very similar. But some times scores are different as much as 10~20kcal/mol even if each job was run in all the same parameters but the random seed.
How do you decided which conformation is best? And When are you sure that you obtained global mimimum or quite reliable result?
Thank you for your detailed solution. And have a nice day!
Graduate student, Computational biology Laboratory
Seoul National University, Korea.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Dock-fans