On Thu, Nov 25, 2010, sunita gupta wrote:
> Hello Everyone
>
> I am calculating relative binding affinity of two ligands using TI. In
> vdw_decoupling step (protein medium). Till heating everything went fine, but
> I am facing this error at the beginning of the equilibration "vlimit
> exceeded for step Coordinate resetting (SHAKE) cannot be accomplished,
> deviation is too large".
If it happens right at the beginning of equilibration, you should check that
nothing has changed between the last few steps of the previous run, and the
equilibration. Set dt=0.001, ntpr=1 and run just a few steps. Make sure the
energies are essentially the same as they were in the previous stage.
I note that you seem to be changing ntr from 1 to 0 at this stage, and that
you had rather tight restraints at the end of the heating stage. You might
want to try an intermediate step with a restraint weight of 0.5 or so. If you
had a high restraint energy at the end of the heating step, and then suddenly
removed it, there could be strong forces that lead to SHAKE failures. It's
possible that this may all go away if you use a more conservative value for
dt.
This is also the point at which you change from ntp=0 to ntp=1. Run some
short tests to see if that is influencing the results. You should also check
if this is related to value of clambda you are using.
...good luck...dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Nov 26 2010 - 07:00:04 PST
