[gmx-users] RE: RE : FEP : separating components of dgdl

[Berk Hess]

> From: gmx3 at hotmail.com
> To: gmx-users at gromacs.org
> Date: Thu, 31 Jan 2008 15:56:01 +0100
> Subject: [gmx-users] RE: RE : FEP : separating components of dgdl
> 
> 
> 
> 
> ----------------------------------------
> > Date: Thu, 31 Jan 2008 15:29:50 +0100
> > From: michael.bon at cea.fr
> > To: gmx-users at gromacs.org
> > Subject: RE : FEP : separating components of dgdl
> > 
> > Hi,
> > 
> > You were right for the bonds, I forgot that I was using constraints = all-bonds.
> > 
> > Could the missing term be the soft core then ? Or is it included in VdW ?
> > 
> No, those terms are always included.
> 
> > Anyway, I noticed that this problem of hidden extra term happens when I modify something else than a hydrogen. Does it make sense ?
> > For example, I did N->DUM with all the parameters kept the same : every components of dVdl are 0 (VdW, Coulomb, angle, proper dihedral, Ryck.,  LJ 1-4, position restraint, disp. corr. ), but dVdl/dlambda is -6. Any idea ?
> 
> I have no clue.
> But the easy way to find out is to take your system with only N->DUM
> and do two runs of zero steps at two different lambda values and see which energy term
> in the log file has changed.

I just realized that the constraint contribution to dVdl is not printed.
In MD this contribution is actually partially dVdl and partially dEkindl, but Gromacs adds it to dVdl.
But you say that you kept everything the same.
Are you sure your bond/constraint lengths did not change?
You can check this easily with gmxcheck -s1 -ab

Berk.


_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080201/82c34d2c/attachment.html>