[gmx-users] pull_pbcatom1 and pull code in gromacs 4

[Chris Neale]

Hello,

The pull-code .mdp option pull_pbcatom1 is defined in the online manual as:

"The reference atom for the treatment of periodic boundary conditions 
For determining the COM, all atoms are put at their periodic image which 
is closest to pull_pbcatom1. A value of 0 means that the middle atom 
(number wise) is used."

I suppose that this is necessary since mdrun now writes broken 
molecules. However, this is suboptimal for replica exchange umbrella 
sampling (pull = umbrella, pull_geometry = distance) in which one does 
not necessarily want to utilize different periodic images of pull_group0 
and pull_group1.

a) Perhaps I don't understand. Is it possible that although a particular 
image of each pulled group is chosen, the pulling is actually still done 
along the vector between the closest two images?

b) If not, is there a simple code modification to utilize a pbc=nojump 
type definition here during dynamics?

Thanks,
Chris.