CHARMM Force Field weighting factor

LAMMPS LAMMPS Mailing List Mirror
1

In the CHARMM dihedral style, it defines a weighting factor. In the CHARMM22 FF, I do not find this weighting factor listed. In the MacKerell paper referenced in the dihedral style, it says there is no 1-4 weighting…So should the weighting factor be set to 0 or 1?

Thanks for the help.

2

In the CHARMM dihedral style, it defines a weighting factor. In the
CHARMM22 FF, I do not find this weighting factor listed. In the MacKerell
paper referenced in the dihedral style, it says there is no 1-4
weighting...So should the weighting factor be set to 0 or 1?

the 1-4 weighting of the LJ term is atom type dependent.
some have special 1-4 LJ terms in the CHARMM parameter
files, others don't. check out the NONBONDED section.

LAMMPS is different from other codes in that the 1-4 terms
are applied in the dihedral calculation and not during the
nonbonded calculation. you still provide the parameters
with the non-bonded atom types, though.

the "special_bonds charmm" shortcut is there to make
sure that you use the right setting for the non-bonded
calculation.

HTH,
   axel.

3

Thanks Axel for your reply. I believe I understand what you are saying. I’ve read the CHARMM dihedral and pair style section and I’m setting up my input file like so:

pair_style lj/charmm/coul/long 12.0 14.0 14.0
bond_style harmonic
angle_style charmm
dihedral_style charmm
kspace_style pppm 0.0001

Some pairs have parameters like:

pair_coeff 1 1 0.11 3.563

and others have pair parameters like:

pair_coeff 4 4 0.02 4.053 0.01 3.385

All of the dihedrals take the form:

dihedral_coeff 1 0.0000 1 0 x --> not sure if x should be 0 or 1

I’m also setting: special_bonds charmm

In my dihedral_coeff, should I be setting x to 1? I have had it set to 0.

Thanks again for the help.

4

In my dihedral_coeff, should I be setting x to 1? I have had it set to 0.

If you set it to 0 you will get no contribution from
the extra pair coeffs defined for the 1-4 atom pairs
in a dihedral, that you listed. If those extra sigma/epsilons
are present, than presumably you want their contribution,
so use x=1.

Steve

5

Steve,

Thanks for the reply. I started setting it to 1 and my simulations are actually giving me good results. It’s funny how one little number can cause crazy results.

Thanks for the help.