[gmx-users] distance issues with umbrella sampling

Nima Soltani nima.slt at gmail.com
Wed Aug 19 20:32:40 CEST 2015


Nash, Anthony <a.nash at ...> writes:

> 
> As far as I¹ve understood, the absolute distance reported using g_dist
> (note: alternative name in 5+) and the reported harmonic potential between
> two groups using pull code in grompp, doesn¹t always match.
> 
> As such, I some times end up with neighbouring umbrella histograms
> practically sitting on top of one another rather than nicely overlapping
> at the tails. Would anyone with more experience than myself at umbrella
> sampling clarify whether I should keep in both histograms, or remove one
> of the hisograms. 
> 
> Also, if my reaction coordinate is an absolute distance and therefor I
> have set pull_dim = Y Y Y (rather than, say, Z for water through an ion
> channel or Y for TM protein dimerisation), is it appropriate to use the
> pullf.xvg and the window.tpr files in g_wham? Or is this beyond the
> capabilities of g_wham. You see, my reaction coordinate is still distance
> but it was defined via a traversed trajectory over a potential energy
> landscape using meta-dynamics.

The "gmx distance" module of Gromacs ver 5 (I am not sure about other
versions) is the same as the distance that pull code calculates .
all you should do is to use this command:

gmx distance -f traj.xtc -s topol.tpr -n index.ndx -oav AverageDist.xvg
-oxyz distxyz.xvg -select 'com of group "A" plus com of group "B"'

 Z (or x and y) coordinate distance can be found at distxyz.xvg output while
average distance can be found at -oav AverageDist.xvg output.

> Many thanks
> Anthony  


-Nima Soltani
Grad. student of physical chemistry
Sharif Univ. of Tech.
> 
> Dr Anthony Nash
> Department of Chemistry
> University College London
> 




More information about the gromacs.org_gmx-users mailing list