Previous Thread
Next Thread
Print Thread
#7120 06/11/05 10:43 AM
Joined: Oct 2004
Posts: 13
S
sunny Offline OP
Forum Member
OP Offline
Forum Member
S
Joined: Oct 2004
Posts: 13
Dear all,

Are there any easy solutions to use vdW Lennard-Jones 10-5 or 8-4 potential (which might be useful in rigid protein docking)in CHARMM insead of normal 12-6 potential?


Thanks in advance,
Sunny

sunny #7121 06/11/05 03:59 PM
Joined: Sep 2003
Posts: 8,605
Likes: 23
rmv Offline
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 8,605
Likes: 23
No easy solutions; I think some Fortran programming would be needed, and probably re-optimization of the VDW parameters, which were developed explicitly for the 12-6 L-J expression.


Rick Venable
computational chemist

sunny #7122 06/12/05 03:49 AM
Joined: Nov 2003
Posts: 23
Forum Member
Offline
Forum Member
Joined: Nov 2003
Posts: 23
Quote:

Dear all,

Are there any easy solutions to use vdW Lennard-Jones 10-5 or 8-4 potential (which might be useful in rigid protein docking)in CHARMM insead of normal 12-6 potential?


Thanks in advance,
Sunny




I believe CHARMM includes a soft core potential (nbonds.doc) that might be appropriate for docking. Not sure what the form is though.

Deepak


"Violence is the last refuge of the incompetent" -- Salvor Hardin Accelrys CHARMm User Group Meetings
dsingh #7123 06/12/05 06:15 PM
Joined: Sep 2003
Posts: 8,605
Likes: 23
rmv Offline
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 8,605
Likes: 23
By default, the soft core potential modifies (softens) both the standard 12-6 VDW term and the ELEC term as well. The documentation in nbonds.doc is especially sparse and cryptic (even for CHARMM), so I'd hesitate to try it w/o at least looking for testcases in the CHARMM distribution and perhaps browsing the source code. Perhaps some kind stranger will post some working examples in the Script Archive.

For docking problems, you may also be interested in the routines described in grid.doc and galgor.doc


Rick Venable
computational chemist

rmv #7124 06/12/05 09:43 PM
Joined: Oct 2004
Posts: 13
S
sunny Offline OP
Forum Member
OP Offline
Forum Member
S
Joined: Oct 2004
Posts: 13
Thanks a lot.

Probably modify the code to get 10-5 potential is not difficult. But which parameters should I re-optimized?

In my case, I will use CHARMM to optimize docked poses (minimization). Then 2 scaling factors will be used to scale down the vdW and elec energy value in order to be compatible to binding energy (the scaling factors are from a training set with experimental data)

I guess the re-optimization of vdW parameters will be not important, since I can adjust the scaling factor, right?

sunny #7125 06/12/05 10:07 PM
Joined: Sep 2003
Posts: 8,605
Likes: 23
rmv Offline
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 8,605
Likes: 23
For that sort of application, it's probaby not necessary to worry about re-optimization. A bigger problem is equating binding energy to scaled VDW+ELEC, as it assumes that the TdS contribution to binding free energy is zero (or equivalent for different molecules).


Rick Venable
computational chemist

sunny #7126 06/13/05 01:19 AM
Joined: Nov 2003
Posts: 23
Forum Member
Offline
Forum Member
Joined: Nov 2003
Posts: 23
Quote:

Thanks a lot.

Probably modify the code to get 10-5 potential is not difficult. But which parameters should I re-optimized?

In my case, I will use CHARMM to optimize docked poses (minimization). Then 2 scaling factors will be used to scale down the vdW and elec energy value in order to be compatible to binding energy (the scaling factors are from a training set with experimental data)

I guess the re-optimization of vdW parameters will be not important, since I can adjust the scaling factor, right?




From what I can gather you are doing something like LIE (although without dynamics). If you are starting off with a docked pose, I think the keys are making sure you have reliable parameters for your ligand(s) and that you are able to get appropriate scaling factors from your training set.

Deepak


"Violence is the last refuge of the incompetent" -- Salvor Hardin Accelrys CHARMm User Group Meetings
dsingh #7127 06/13/05 07:17 AM
Joined: Oct 2004
Posts: 13
S
sunny Offline OP
Forum Member
OP Offline
Forum Member
S
Joined: Oct 2004
Posts: 13
To dsingh,

Yes, you are right. We are doing something like LIE but skip the MD part. The parameters of small ligands are really annoying. We are now using MSI parameters, but still more than 20% compounds missing parameters. What else parameters can be used?

To Rick Venable,
TdS is a big problem in docking. In our application, we assume the TdS contribution are same for similar scaffold ligands, ideally, they are cancelled out.


Moderated by  BRBrooks, lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5
(Release build 20201027)
Responsive Width:

PHP: 7.3.31-1~deb10u1 Page Time: 0.012s Queries: 30 (0.008s) Memory: 0.7626 MB (Peak: 0.8297 MB) Data Comp: Off Server Time: 2022-09-27 02:21:04 UTC
Valid HTML 5 and Valid CSS