Re: targetCenters in Umbrella Sampling

From: Vermaas, Joshua (
Date: Wed Aug 24 2016 - 18:04:31 CDT

Hi Olga,

centers is the value to which the collective variable is restrained by
the harmonic potential. This will change linearly (by default anyway) to
the value specified by targetCenters over targetNumSteps steps. In
effect, this performs a steered molecular dynamics calculation if
targetCenters != centers. If targetCenters is left unspecified, it acts
like it is equal to centers, and is what I use for my own umbrella
sampling calculations. The reason it only runs with small values for the
target is probably because you are still restraining the center to be
VERY far from where the system starts, which means huge velocities if
the force gets to act over more than one timestep. You should see this
in your traj file, where the reported colvar value is going to be much
smaller than 100.


On 08/24/2016 04:54 PM, Olya Kravchenko wrote:
> Hi all,
> just another question about umbrella sampling calculations for ion
> permeation and colvar input file.
> The harmonic restraint part for my input file looks like this now:
> harmonic {
> colvars test
> centers 134.0
> forceConstant 5.0
> targetCenters 108.0
> targetNumSteps 1
> }
> I would like to measure PMF for the ion moving between 134 and 108. Do
> I understand it correctly that I have to use 108 for targetCenters?
> When I use 108, namd quits. But when I use smaller numbers (up to 54)
> it runs normally. Why would that be the case?
> Olga

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:22:25 CST