Re: MD simulation Patch grid errors

From: Jeffrey Potoff (jpotoffx_at_gmail.com)
Date: Tue Jan 07 2014 - 18:03:29 CST

On 1/7/2014 6:33 PM, Jiann George Pazhamalil wrote:
> Dear NAMD users,
>
> I am a student and extremely new at this process and I am trying to
> run a protein and membrane simulation where I am receiving the error
> "FATAL ERROR: Periodic cell has become too small for original patch
> grid!" I am aware that this is a common problem and I have been
> looking at the archives however increasing the margin or restarting
> the run is not helpful and I am still receiving this error and with a
> membrane, I can't switch off flexiblecell. Energy minimization and
> equilibration is running fine but the problem is with the MD run which
> is unique from the previous problems seen on the list. I am a little
> unsure about the surfaceTensionTarget value I am using. I have a POPC
> bilayer and I'm using 17 dyn/cm, could that possibly be the reason of
> the error?
>
With a membrane simulation, there many things that can go wrong that
will result in "patch grid" error, such as

1. Not enough water in the system (density is too low).
2. Initial membrane structure is far from equilibrium.
3. Poor lipid model. Older lipid force fields had the tendency to
"condense" when used in constant pressure simulations.

Some solutions:
1. Use a one of the modern lipid force fields, like CHARMM 36, which
does not require the application of a surface tension to produce the
correct area per lipid.
2. Equilibrate for ~2-4 ns using NVT, then switch to NPT.
3. Use an anisotropic pressure move. In NAMD this is done by setting
"useconstantratio" to "yes".
4. Make sure your initial configuration is reasonable. CHARMMGUI is
pretty good for the generation of membrane structures.
5. If all else fails, try using a smaller timestep (0.5-1ns) for 0.1-1
ns to get the system to calm down.
6. Set your DCD and restart output to something small, say 100 steps
and run a short simulation ( < 10,000 steps). Sometimes you can get
NAMD to drop a restart file that you can bootstrap off to get a stable
simulation going.

-- 
======================================================================
Jeffrey J. Potoff                         jpotoff_at_wayne.edu
Associate Professor and Director of Early Engineering Programs
Department of Chemical Engineering and Materials Science
Wayne State University			  5050 Anthony Wayne Dr
Detroit, MI 48202
http://potoff1.eng.wayne.edu
======================================================================

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:02 CST