[gmx-users] problem with NVE ensemble simulation with Martini lipid

[Mark Abraham]

On 12/04/2010 5:01 AM, Sanku M wrote:
> Hi,
> I am having a problem with running a NVE ( micro-canonical) ensemble
> using a system of 128 Martini DPPC lipids. I started the simulation with
> a 128 lipid bilayer with Martini water which was pre-equilibrated in NPT
> ensemble ( using berendsen temperature coupling at 325 K( separate
> coupling for lipid and water) and pressure coupling at 1 atm for 500 ns).
> But, when I started running the NVE simulation by switching off pressure
> coupling and temperature coupling, the temperature started going down
> from 325 to almost 2 K very quickly which looks like very unphysical.
> However, total energy remains conserved . But the temperature is
> becoming a concern as it is going down to a 0 K. I am not sure whether I
> am doing some mistake in the mdp parameter. Any help will be appreciated.

I don't know anything about running Martini simulations but I could see 
that the following are not consistent with running a pre-equilibrated 
system:

> ; GENERATE VELOCITIES FOR STARTUP RUN
> gen_vel = yes
> gen_temp = 325
> gen_seed = 473529

You want to keep the old velocities, not make new ones. See some 
tutorial material or the webpage about run continuations for 
instructions here.

I don't know if this is the cause of your problem, however.

Mark