Re: Problem with Alchemical FEP

From: Jawahar Neelankatan (
Date: Wed Aug 09 2006 - 19:56:19 CDT

Hi Michel and Jerome,
 Thanks for the input. Yes, the flexiblecell option was always disabled while I was still getting the errors.
The box contains 1089 water (TIP3P) molecules and it was generated by using VMD Solvate 1.2 to solvate the ions. The average unit cell length of the box is about 32 Angstrom.
 My solution to the problem was to equilibrate the box generated by VMD Solvate. I did a short 500 ps equilibration of the box (regular simulation, not FEP). Then I extracted the last frame of the dcd file generated from the equilibration and used that as my starting pdb file. It works now. Thanks.


Jerome Henin <> wrote: Jawahar,

First, I strongly back Michel's suggestion of enforcing isotropic cell
fluctuations. Actually, that kind of system should almost always be run in a
cubic cell, as a compromise between efficiency and simplicity.

My other question is: how big is your water box?

By the way, did you find the solution to your xsc file problem?


On Monday 07 August 2006 12:38, L. Michel Espinoza-Fonseca wrote:
> Just out of curiosity, are you using an anisotropic periodic cell? You
> can take a look to your input file and see if the option
> "flexiblecell" is "on". Unless you're simulating a system containing a
> membrane, you should disable this option. It is weird that you're
> setting such a big patch and you're still getting this problem. Maybe
> you should pay attention to the preparation stage of your system. This
> will be my guess.
> Michel
> 2006/8/7, Jawahar Neelankatan :
> > Hi all,
> > I'm trying to do an alchemical FEP simulation of an ion in TIP3P water.
> > I'm using CHARMM parameters but I get the following error at the
> > beginning of the simulation:
> >
> > FATAL ERROR: Periodic cell has become too small for original patch grid!
> >
> > I've tried pretty much everything, from increasing the margin (up to 70)
> > to tweaking the PMEGrid sizes. It always gives the same error. The odd
> > thing is that if I
> > delete the portion of the configuration file devoted to FEP simulation,
> > and just run the simulation like a normal simulation, it works fine. Does
> > anyone know how to get around this problem ?
> >
> > -Jawahar
> >
> >
> > ________________________________
> > Do you Yahoo!?
> > Next-gen email? Have it all with the all-new Yahoo! Mail Beta.

Do you Yahoo!?
 Next-gen email? Have it all with the all-new Yahoo! Mail Beta.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:27 CST