[AMBER] different energy values from mdout and cpptraj

From: Batuhan Kav <bkav13.ku.edu.tr>
Date: Tue, 7 May 2019 14:24:18 +0200

Dear All,


After revisiting some old simulations, I would like to calculate/obtain
the total energy of the system as a function of time. The simulation
setup consists of two monomers in explicit solvent. The main issue is
that I did not save the coordinates of the water molecules in the
trajectory, so I cannot run cpptraj energy function to obtain the total
energy (energy with the solvent). At this point, I tried to consult to
the mdout file as it contains the energies in every some step.

As the saved trajectory does not contain all the atoms, I wanted to
compare the dihedral energies from mdout file and cpptraj/energy
function. That was a consistency check for me because regardless of the
contribution from the solvent, I think the dihedral energies should be
the same. However, what I realized is that, the energy terms saved in
mdout file do not match with the energy values I obtained after running
cpptraj/energy function. Although I save the trajectory more often than
I save the mdout file, none of the energy values calculated from cpptraj
match with the ones in the mdout file. I should add that if I set ntpr=1
and ntwx=1, then both cpptraj and mdout values match but for any other
combination of ntpr and ntwx I cannot reproduce the energies in mdout
file with cpptraj/energy command.

I would like to ask, if the energy values reported in mdout file
corresponds to averages over certain steps. If not, what might I be
doing wrong?

The mdin file is as follows:

Production
&cntrl
imin = 0,
ntwprt = 148,
irest = 1, ntx = 5,
ntb = 2,
ntc = 2,
ntt = 3,
gamma_ln = 3,
ig = -1, !
ioutfm = 1,
ntp = 1,
barostat = 2, pres0 = 1.0,
dt = 0.002,
nstlim = 1000000000,
temp0 = 298, tempi = 298,
cut = 10,
ntpr = 100000,
ntwr = 100000,
ntwx = 5000,  
ntxo = 2,
/

I am using cpptraj V18.01. The simulations were performed with Amber 16
using pmemd.cuda.

Thanks for any possible suggestions.

Batuhan


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 07 2019 - 05:30:02 PDT
Custom Search