Re: Atoms moving too fast OR constraint failure in RATTLE algorithm

From: Diego Alejandro Vargas (
Date: Wed Nov 05 2008 - 23:17:56 CST

Thank you for your quick reply, I appreciate it.

I moved the langevinPistonTemp line into the loop. This parameter has to be
positive, but it has to be the same as whatever parameter is controlling the
temperature (I pressume langevinTemp). So they cannot be zero, therefore I
started it at 1K before the first iteration of the loop.

With respect to letting the temperature increase cycles last longer: I will
keep it in mind for later, for now it does not make a difference, since as
you assumed it crashes immediately. After moving the pressure control line
into the loop it actually performs the first step (step 0), but atom
velocities are in the ~10^35 order of magnitude. By changing once more the
margin, nonbondedfreq, fullelectfrequency, and timestep parameters these are
reduced, but still too high (~10^5): Also, this leads to many more atoms to
be moving above the allowed velocity.

Yet in some cases it gives me the same error message of velocities being
high for atoms in which velocity is within the limit (e.g. Atom 37843
velocity is 44302.7 -21379.8 12750.8 (limit is 50000)). Am I misinterpreting

With respect to the XST file, it shows no change in the minimization
simulation. And in the equilibration simulation it doesn't go past the step
0. So this does not say much.

Finally, I don't completely understand the beta-factor. What would I be
monitoring for, beta factor in resulting dcd file from minimization? It
would display the stability of the system (or atoms) would it not? When
coloring the molecule based on the beta factor through the simulation it is
uniformly colored.

What I did notice by doing this, and what is a bigger problem is that the
protein now sticks out of the water box. The water molecules are more
concentrated and don't cover the water box. Does this make sense? In the psf
and pdb files from the minimization simulation the water box is bigger (by
20 Angstrom), but it doesn't show this in the beginning of the trajectory
dcd file.

Do you have any additional input?

Thank you,

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:21:27 CST