[gmx-users] Distance for window spacing in US?

Justin Lemkul jalemkul at vt.edu
Wed Feb 20 14:16:39 CET 2013

On 2/20/13 1:14 AM, Yun Shi wrote:
> Hi all,
> For the purpose of calculating deltaG from umbrella sampling, I have
> first pulled my ligand away from the enzyme active site along an
> arbitrary direction (say x axis), setting the enzyme as pull_group0.
> But I found the ligand was moving in the yz plane while pulling along
> x, which I guess is not uncommon. Then I wonder which distance to use
> a reference for the definition of umbrella sampling windows. Would
> that be better to stick to the x component of protein-ligand distance,
> which I can easily read from pullx.xvg file while ignoring the y and z
> components? Or is it still better to use g_dist to calculate the
> absolute protein-ligand COM distance and use it for spacing reference?

If you have motion in all three spatial dimensions, I would think that ignoring 
deviations in any of them is inappropriate.  Single-dimension reaction 
coordinates only make sense if the desired path is linear along one axis.  If 
you're pulling something out of a binding cavity or channel that is not exactly 
linear, that complicates matters.  There were some nice explanations of 
practical considerations for such simulations a few months ago, so do check the 



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