 electrostatic energy calculation
|
Joined: Sep 2004
Posts: 29
Forum Member
|
OP
Forum Member
Joined: Sep 2004
Posts: 29 |
Dear All
I try to write a program that calculates different charmm energy contributions to the interaction enenrgy between two selection using charmm19 force-field. I managed to get vdw contributions correctly (the same as calculated from charmm within the roundoff errors), but failed for electrostatic. Since I want to have rdie I use the equation 15 of the original charmm paper:
Eelec = qi*qj/(4PI*eps'0*rij^2)
I assume eps'0 = 1 (since EPS = 1.0 in my charmm.out file)
but the values I got from my program for a pair of atoms (if using ATOM) or group of atoms (if GROUP) are 10^4 too small. What equation should I use then? What's teh value of eps'0?
By the way, the the ratio between my and charmm elec energy values are around 4172.0 (vary between 4171.5 and 4173.5), so I can use this value as a proportionality factor. But could you tell me what it is?
Thank you in advance for your help
marcin
Last edited by marcin; 11/01/06 02:41 PM.
|
|
|
 Re: electrostatic energy calculation
|
Joined: Sep 2003
Posts: 4,803 Likes: 2
Forum Member
|
|
Forum Member
Joined: Sep 2003
Posts: 4,803 Likes: 2 |
This is just Coulomb's law and you have to use appropriate units and values for everything, including eps0 (the dielctric permittivity of vacuum, which is not 1 in most unit systems). With the units used in CHARMM (AKMA, see usage.doc) you get
E=332*qi*qj/rij
The charge and distance units are |e| and A.
Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
|
|
|
PEPtide INstability IN VACUUM...
|
Joined: May 2006
Posts: 11
Forum Member
|
|
Forum Member
Joined: May 2006
Posts: 11 |
Can anybody present a precise idea why charmm (c32b2 with CMAP correction) in vacuo transmembrane peptide simulation fails to reproduce stable folded peptides ?
I observe a strong propensity for turn and coil (in STRIDE assignment of a 2.5 ns trajectory) rather than alpha helices which I would expect according to established theories.
Is this behaviour due to FF optimization for condensed phase? or a e-stat. problem (with "exaggerated" partial charges) ? many thanks for helpful suggestions, David Minde
|
|
|
Re: PEPtide INstability IN VACUUM...
|
Joined: Sep 2003
Posts: 8,524 Likes: 2
Forum Member
|
Forum Member
Joined: Sep 2003
Posts: 8,524 Likes: 2 |
This should probably be a new post; it is unrelated to the initial topic of this thread.
Vacuum may not always be a good model for a lipid bilayer.
Rick Venable
computational chemist
|
|
|
Re: PEPtide INstability IN VACUUM...
|
Joined: May 2006
Posts: 11
Forum Member
|
|
Forum Member
Joined: May 2006
Posts: 11 |
Quote:
This should probably be a new post; it is unrelated to the initial topic of this thread.
Vacuum may not always be a good model for a lipid bilayer.
Yes&Yes. Unfortunately, I was not able to find the button for setting up new posts in the forum. Is it possible that it is not displayed with some operating systems (e.g. FC 5_64bit)?
I need vacuum simulations just as a "tiny piece" of evidence and for comparison with special mass spectrometry measurements. No doubt, some much better "membrane" models have been developped so far.
many thanks,
David
|
|
|
Re: PEPtide INstability IN VACUUM...
|
Joined: Sep 2003
Posts: 8,524 Likes: 2
Forum Member
|
|
Forum Member
Joined: Sep 2003
Posts: 8,524 Likes: 2 |
The Post button is shown in the forum thread index, at the top of the page. It is not shown when viewing a thread.
My bias is toward explicit solvation, despite the computational expense, although I've seen some reports that dielectric slab models can be useful.
Rick Venable
computational chemist
|
|
|
Re: PEPtide INstability IN VACUUM...
|
Joined: May 2006
Posts: 11
Forum Member
|
|
Forum Member
Joined: May 2006
Posts: 11 |
thanks for your help.
(I hope to continue to learn from both explicit/implicit models)
|
|
|
|
|
