It's the force on the spring connecting the two groups.

So, then, it's what I said. In any case, I must say that the probe along Z was more representative of the reality: a hemispherical slice probes a circle near the surface of the membrane, which has nothing to do with the pore, but naturally produces a barrier. I wish there was a way to probe a spherical symmetry within a given solid angle. If pullf/pullx data had any angular info, we could script our own implementation of WHAM, but alas.


code would clarify, because the result of wham now looks rather strange -- a long (1.5 nm!) hydration barrier for the ion. Makes little physical sense and very different from the bias along Z case.

I am guessing that the sampling of 1 ns per hemispherical slice is far from sufficient to probe that space... Just looked at one of the trajectories there -- the ion is nowhere near moving around the entire slice.


On 3/13/2018 10:27 AM, Justin Lemkul wrote:

On 3/13/18 10:32 AM, Alex wrote:
It appears that things have improved... I am just glancing and pullf and pullx data from the first sample config and both contain only one value, even though we set dim to 3D. Does this make sense?

I haven't used the 2018 pull code yet. Presumably this is correct, just the distance rather than all its components, though that would be a change from previous versions.




On 3/13/2018 8:18 AM, Justin Lemkul wrote:

On 3/13/18 10:17 AM, Alex wrote:
You may have a situation that requires manual assignment of pull_group1_pbcatom1, as the automatic choice of

Sorry, that should be pull_group1_pbcatom. Must remind myself not to email before coffee...

Did you mean pull_group2_pbcatom? In my case:

pull_group1_name        = K
pull_group2_name        = CNT

The language for 5.1 pull is (you be the judge on how clear it is):


   (0) The reference atom for the treatment of periodic boundary
   conditions inside the group (this has no effect on the treatment of
   the pbc between groups). This option is only important when the
   diameter of the pull group is larger than half the shortest box
   vector. For determining the COM, all atoms in the group are put at
   their periodic image which is closest topull-group1-pbcatom
<>.    A value of 0 means that the middle atom (number wise) is used. This    parameter is not used withpull-group1-geometrycylinder. A value of
   -1 turns on cosine weighting, which is useful for a group of
   molecules in a periodic system,/e.g./a water slab (see Engin et al.
   J. Chem. Phys. B 2010).

Indeed, my group2 is a membrane that is wider than the shortest half-box. I am assuming it would be a single atom belonging to the pore region...

Yes, pull_group2_pbcatom, not 1.




Gromacs Users mailing list

* Please search the archive at before posting!

* Can't post? Read

* For (un)subscribe requests visit or send a 
mail to

Reply via email to