Hi all,
As a prelude to parameterizing some ions, I am trying to calculate the
solvation free energy of K+ (potassium) using the various TI bits in
sander. I have no problem "uncharging" potassium from +1 to 0 --
everything works fine and I get what seems to be a very reasonable
deltaG for this step (~72.6 kcal/mol). When I use the softcore
potential facility to make the neutral potassium atom turn into
nothing (i.e., disappear), however, I run into a weird problem...
My initial minimization and NVT heating steps finish fine and I don't
see anything out of the ordinary in the output or restart files for
those parts. However, when I try to perform an NPT/density
equilibration, the simulation hang just after reading the restart
coordinates/before starting dynamics without dying or giving any sort
of error (see link below). I was careful to be sure that the
\lambda=0 system was the neutral potassium + water box and that the
\lambda=1 system was just the water box (and that these were properly
generated by tleap). Likewise, I put the potassium atom in the scmask
for \lambda=0 system and left the scmask for the \lambda=1 system
blank. Overall I think that the system setup is OK, so I'm at a bit
of a loss to figure out where to start the troubleshooting.
In the hopes that someone might spot something obvious, I've posted
all my files to: http://www.ocf.berkeley.edu/~psn/potassium/ The
naming convention is pretty simple -- K0_wat is for the \lambda=0
system, K0_watonly is for the \lambda=1 system. All the mdin scripts,
group files, etc. are generated by the bash script setup_vdw.sh
(currently set to generate scripts just for one lambda value). The
last written output of the NPT equilibration is here: http://www.ocf.berkeley.edu/~psn/potassium/output/K0_wat_vdw_equi_v0_l1.out
Many thanks in advance,
Paul
P.S. If all else fails, I realize that I could go the dummy atom route
and probably use the "old" (i.e., non-softcore) TI routines without
issue, but I figured it was worth trying to fix this first.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 18 2010 - 18:30:03 PDT
