AMBER Archive (2009)

Subject: [AMBER] Steered MD.....SHAKE error!!

From: Hlengisizwe Ndlovu (phy5hn_at_leeds.ac.uk)
Date: Fri Jul 17 2009 - 03:50:11 CDT


Hello,

I am currently carring out SMD siumaltions using the nscu_SMD function
of amber10. Prior
to this, my systems have gone through minimiastion and equilbration
steps with no
problems. I have even run seperate 20ns of MD without any problems
with sensible
output---just to be sure its not a 'minimisation' issue.

However, when carring out the SMD siumaltions on which i use 3pairs of
atoms for constant
velocity pulling all seems to work fine until late in the simulation
when i get the error
message below:

vlimit exceeded for step 80727; vmax = 33.2222

      Coordinate resetting (SHAKE) cannot be accomplished,
      deviation is too large
      NITER, NIT, LL, I and J are : 0 5 745 1522 1523

      Note: This is usually a symptom of some deeper
      problem with the energetics of the system.

The odd thing is that sometimes the SMD simulations do go to
completion without the error
and sometimes they dont...depending on pulling velocity used!!!

Does this have anything to do with the 'restraints' imposed by
nscu_smd on the pair of
atoms etc.....

at a loss as to the cause as i would have expected this type of error
during minimisation
or very early in the MD.

cheers

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber