- This topic has 1 reply, 2 voices, and was last updated 14 years, 3 months ago by Anonymous.
July 27, 2009 at 8:38 pm #527Anonymous
We are trying to compare two possible docking sites between partners A and B. The problem is that we are unsure on how to compare the scores.
If we prepack the two sites independently using exactly the same structures for A and B but a different relative orientation, and using the same random_seed, even the minimized_away entry in the fasc file have quite different scores. So, how to compare the scores from two perturbation runs on the two sites ? By looking at the difference between the score and the energy of the minimized_away on the two sites ? (we think this could represent the interaction energy)
Another option we thought of, is to prepack one orientation, take the .ppk.pdb and manually reorient the two partners as in the second site.
Any suggestion ?
September 1, 2009 at 9:16 pm #4169Anonymous
The scores following prepacking are not what you are looking for because in prepacking the sidechain conformations are optimized while the partners are far apart from each other, and such are irrelevant to a given interface.
The best way to compare two potential binding interfaces is by generating a small number of decoys (10-20) at each interface using only high-resolution refinement. To do this, use a starting structure representing the respective interfaces, and do standard docking with the -docking_local_refine flag. Once all decoys are completed, compare the lowest energy decoy generated from each interface.
An alternate way is to do an entire local perturbation docking run using each interface as the starting structure. You will need around 700-1000 decoys, and use the flags -dock_pert 3 8 8 and -spin. Using this method you can look at not only the lowest energy structures at each interface but also the energy funnel at each interface. Look up FunHunt for more information about comparing docking funnels.
- You must be logged in to reply to this topic.