From: Chris Moth <>
Date: Thu, 5 Oct 2017 07:16:11 -0500

A nice thing about any computer software is that the source of a
calculated number can always be found by debugging the code to the point
that the number is generated.

Build a three-atom one molecule simulation cell that demonstrates the
1134kJ/mol.  Then, set breakpoints int the code to reveal the various
calculations that arrive at 1134kJ.

The precise way you do this will vary depending on the programming
language of the code.  But, "print x" is easy code to add.  If you have
a debug mechanism where you can set a "breakpoint" at code startup, that
is even better.

On 10/5/2017 4:44 AM, Hossein Geraili wrote:
> Dear Amber users,
> I am not using the Amber software, but I could not find a place that people
> would know about the right combination of Amber and MD. Please if you can
> comment help me with this. I am using Lammps for my simulation.
> All the bonded energy (bond, dihedral, improper) are equal to article's
> value, but the angle energy value is so big(1134 kJ/mol but 53 kJ/mol in
> the article). What can be wrong? I checked the angle parameters several
> times as well as the commands, but none of my theoretical friends found out
> what is wrong. If other energies were wrong too, it would not be a problem.
> Also, I checked the angle parameters from other sources( Amber force field
> article for my molecule) and they are almost the same. Moreover, before
> this Amber ff, I used Charmm force field and I did not get right values for
> density, and recently I checked it and saw its angle energy is also so big.
> Because of these big angle energy values total and potential energy in
> both( Charmm and Amber) simulations are big positive values, but in both
> articles, they are negative. Is there something special about angle
> charmm and amber force field?
> I am using Amber force field. Maybe this will not be helpful, but this is
> my input in case that somebody can use these to figure out the problem:
> units real
> boundary p p p
> atom_style full
> bond_style harmonic
> angle_style charmm
> dihedral_style charmm
> improper_style cvff
> pair_style lj/charmm/coul/long 13 15
> pair_modify mix arithmetic tail yes
> kspace_style ewald 1.0e-4
> read_data
> neighbor 2.0 bin
> neigh_modify delay 2 every 1 check yes
> special_bonds amber
> minimize 1.0e-10 1.0e-12 10000 10000
> velocity all create 313 4928459 rot yes dist gaussian
> timestep 2.0
> run_style respa 4 2 2 2 pair 1 kspace 4
> fix 1 all nve
> run 10000
> unfix 1
> fix 3 all npt temp 313 313 30.0 iso 0.9671848 0.9671848 700.00
> run 3000000
> Please if you have any comment, let me know.
> Thanks in advance

AMBER mailing list
Received on Thu Oct 05 2017 - 05:30:03 PDT
Custom Search