[gmx-users] energy_increase_NVE_run

[Justin A. Lemkul]

Durba Roy wrote:
> Hi Gromacs Users,
> 
> I am trying to run a NVE simulation of a single dye molecule jumping 
> from state S0 to state S1. I have starting coordinates and velocities 
> from a previous NPT run of the same system at S0, however, the charges 
> are new, that of S1. I didnot generate any Maxwell-velocity for the NVE 
> run and followed the instructions that rlist should be ~0.3 nm larger 
> than rcoulomb and rvdw.  I also increased Lincs_iteration value and the 
> table extension.  Please note that I am using a tabulated interaction 
> potential for a particular dihedral.
> 

Simulating under a different ensemble, then changing simulation parameters (like 
cutoffs) does not bode well for simulation stability.  Note, too, that there are 
other factors that can influence the stability of an NVE run:

http://www.gromacs.org/Documentation/Terminology/NVE

> The problem is, an increasing total energy during the NVE run and the 
> molecule collapses followed by segmentation fault. However, the S0-S1 
> jump sim runs fine in NPT ensemble.
> 
> Also, when I use null potential for the dihedral the NVE job runs fine.
> Kind of confused and will really appreciate your help.
> 

If your system is stable without the tabulated potential, then breaks down when 
you use it, it seems pretty clear that however you're treating this dihedral is 
causing your simulation to break.  Without more specific details (your full .mdp 
file, the parameters you're using for the dihedral, etc), it will be hard to 
give you any further help.

-Justin

> Thanks,
> Durba
> 

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================