On May 7, 2014, at 10:03 AM, Lara rajam <lara.4884.gmail.com> wrote:
> Dear Amber
>
> I am doing a protein ligand complex simulation
>
> I am trying it to do in NPT ensemble. I restrained and did the constant
> temperature equilibration.
> I kept the restrain force constant as 100 for the solute molecule .
This is an incredibly strong restraint. Consider using a value between 1 and 2 orders of magnitude smaller. I typically use restraint force constants between 1.0 and 10.0 when running dynamics (note that 10 kcal/mol/AA^2 is a strong restraint). Larger force constants make integration errors more likely (or require shorter time steps).
> when I did the same for the constant pressure in the second step the
> equilibration is stopped in the middle by saying the error
>
> vlimit exceeded for step 79248; vmax = 29.6169
> vlimit exceeded for step 79254; vmax = 539.7187
>
> Coordinate resetting (SHAKE) cannot be accomplished,
> deviation is too large
> NITER, NIT, LL, I and J are : 0 0 864 1810 1811
This message contains some potentially useful information -- atoms 1810 and 1811 are the constrained pair that seem to be causing problems.
> Note: This is usually a symptom of some deeper
> problem with the energetics of the system.
_Something_ is causing a very large force on one of the atoms in your system. It could be the position restraints if one of the atoms got far away from its reference position (this can happen from integration errors or if the reference structure is significantly different from the starting structure of the simulation).
Alternatively, your starting conformation is partially strained, leading to strong forces arising from van der Waals or electrostatic repulsion, for instance. If you have been restraining the entire solute (including side chains and hydrogens) with a strong restraint, then you could have an unstable region where strong steric or electrostatic clashes are competing with the overly strong restraint forces, ultimately leading to integration errors.
This is why I usually use force constants on the order of 1 to 2 kcal/mol/A^2 when equilibrating (and I also only restrain the backbone, letting the sidechains move freely). This way my restraints are not overstabilizing a locally strained conformation.
At the end of the day, the Amber potential is fairly simple -- there are not many possible explanations for how forces large enough to cause SHAKE errors can occur. Finding the actual cause will take a little bit of investigation -- visualizing your system always helps, as can the "checkoverlaps" command in cpptraj, which should help identify possible steric clashes.
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 07 2014 - 08:00:04 PDT
