Protein hydrogens causing "Atoms too fast" error

From: Christopher Gillespie (
Date: Mon Dec 03 2007 - 18:03:59 CST

Hello all,

I am trying to hydrate a protein with TIP3P waters using a box of pre-
equilibrated waters that fails at the beginning of equilibration. I
have followed the tutorials and user guides to remove any overlapping
waters from the protein's surface. I minimize the combined file for
anywhere from 500 to 20000 steps and have found that the protein's
hydrogens always minimize such that they turn away from the solvent
and back to the protein's heavy atom, such as N or O. Therefore, the
hydrogen and N or O atoms are almost overlapping. I found these by
using the identified atoms that are moving too fast. I tried both an
immediate jump to 298 and a slow heating over 10ps, both failing at
first step. Now, the protein's heavy atoms, i.e. not hydrogen, are
fixed since my work is based on the crystal structure and the final
state in the crystal, so moving the heavy atoms during minimization
will not help. I have changed the timestep down to 0.1 which does
run, but running at those conditions is too computationally expensive,
but it still fails anywhere after 100 steps.

Things I have checked:

No atoms have coordinates of 0,0,0.
Periodic cell dimensions (yes they are large enough)
PME grid (CellBasis 60x60x60, Grid 64x64x64)
"Correcting" the hydrogen coordinates to pre-minimization location,
only the ones ID'd as moving too fast
        -Result: different protein hydrogens giving error

This is only the first step of putting two proteins in contact, so any
help with this is appreciated.



This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:45:39 CST