[gmx-users] deltaG from PMF

Justin Lemkul jalemkul at vt.edu
Tue Aug 21 16:27:08 CEST 2012

On 8/21/12 10:23 AM, Steven Neumann wrote:
> On Tue, Aug 21, 2012 at 3:07 PM, Justin Lemkul <jalemkul at vt.edu> wrote:
>> On 8/21/12 9:36 AM, Steven Neumann wrote:
>>> Dear Gmx Users,
>>> I got confused about reading free energy difference from PMF curve.
>>> It is the difference between maximum value on PMF curve (plateau -
>>> final state) and the starting point corresponding to minima.
>>> So e.g. my curve starts at 0 [kcal/mol] going staright to minima of -2
>>> kcal/mol and then increase obtaining plateau of 14 kcal/mol. DeltaG =
>>> -16 kcal/mol or -12 kcal/mol ?
>> It seems that your system has discovered a more energetically favorable
>> location where the minimum is at -2 kcal/mol.  It should not be assumed that
>> whatever arbitrary "start" point we choose is the energy minimum.  In your
>> case, it is not.  DeltaG is the difference between final and initial states.
>> Only you know what these states are.
>> -Justin
> Thank you. Ok - mu curve starts at 0 kcal/mol corresponding to the
> reaction coordinate of 0.5 nm. going to -2 kcal mol at 0.6 nm then
> increasing as I said before to plateau of 14 kcal/mol.
> My first window (shown by grompp) corresponds to 0.8 nm of the
> reaction coordinate... So the initial state should be at 0.8 nm which
> corresponds to 0.5 kcal/mol or always take the initial state as
> initial of the curve?

That depends on how you've set up your .mdp files.  If you're telling a 
configuration in which the groups are separated by 0.8 nm that they should be 
separated by 0.5 nm instead, you need to allow for some time for that to adjust. 
  The restraint distances are what you tell them to be.  Your initial 
configurations should be as close to that COM separation as possible.

So, the answer is, I don't have an answer, because you didn't provide an .mdp file.



Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list