AMBER Archive (2008)

Subject: AMBER: PMEMD with AMOEBA parms including a lone pair site

From: Margaret Johnson (
Date: Tue Nov 18 2008 - 20:29:04 CST


I am trying to run pmemd using AMOEBA parameters (which I have
previously done succesfully), where the DMSO molecule has an additional
'M' site defined at the lone pair position.
While I have been able to generate the inpcrd and prmtop files using
/amoeba_parm, /after one step in a pmemd NPT simulation, all the
velocity dependent properties are NaN, while all the potential energy
values are fine.

 NSTEP = 1 TIME(PS) = 0.001 TEMP(K) = NaN PRESS
= NaN
 Etot = NaN EKtot = NaN EPtot =
 BOND = 226.9192 ANGLE = 597.6138 DIHED =
 1-4 NB = 550.2788 1-4 EEL = 0.0000 VDWAALS =
 EELEC = -21829.1325 EHBOND = 0.0000 RESTRAINT =
 EPOLZ = -3684.2734
 Dipole convergence: rms = 0.469E-04 iters = 14.00
                                                    Density =

When I run the system using tinker programs, including /dynamic/, the
system propagates succesfully. I have previously generated inpcrd and
prmtop files with AMOEBA parameters (using /amoeba_parm) /for other
systems that ran succesfully with pmemd, the only difference this time
is the M site.
This site has a mass of 1, and bond and angle constraints to maintain
its position, but zero everything else, which are explicitly listed out.
(vdw, torsions, multipoles, polarizability).
I also tried restarting in pmemd using the /.dyn /output from a tinker
dynamic run, which had the same problem as above.
Is there some way to fix this? It doesn't seem like the M site is
defined any differently than any other atoms, except for all the zeroed

Thanks very much,
Maggie Johnson

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)