On Tue, Oct 12, 2010, xuemeiwang1103 wrote:
> Thanks a lot for your comments. There is a metal ion in my protein
> sysytem.It tetracoordinated to three residues as well as one water
> molecule.The number 9223 and 9224 is the H atom in the water
> molecule which bonded to the metal ion. I used the bonded model
> in Amber. The parrel simulation runs well during the heating
> process,but when I run the MD in the NPT,the error happens.
Can you be a little more specific? Are atoms 9223 and 9224 both H atoms in
the same water molecule? Are they in the same residue? Sander should not
be splitting atoms in the same residue up among processors (although it is not
clear whether you are running sander or pmemd. Also which version?)
It is indeed odd that the partition failure happens with NPT and not NVT.
I assume(?) you used the same number of processors for each. Maybe answers
to the questions above will help, but you will probably have to post your
prmtop and coordinate files in order for someone here to try to be able to
reproduce the problem.
...thx...dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Oct 11 2010 - 21:00:02 PDT
