Charges in DPD prm

EMC
1

Hi veld,

I have some remarks regarding generated LAMMPS data files using dpd forcefields.

  • The order of columns is not consistent with the data files generated using atomistic forcefieds (at leat oplsaa). Using atomistic forcefield, the order of column is “id resid type charge x y z” ; whereas for dpd it is “id type x y z resid charge”.

  • Charges are not taken into account in dpd .prm files, or I have missed a point. Charges are always 0 even if I change this specific part of .prm files :

    # Masses
    ITEM	MASS
    # type	mass	name	ncons	charge	comment
    *	2	*	2	8	charged

Masses are taken into account. I found a workaround, i.e. I specify charges directly in esh file :

STM         *[Bs]([Ph][Tmm][Tma+1])([Ph])*, &
        1,STM:2,  1,BUM:2,  1,STE:1, &
        2,STM:1,  2,BUM:1,  2,STE:1

You can find attached .prm and .esh files.
aem.prm (1.6 KB)
sebs_aem_w.esh (877 Bytes)

Best,

2

Hi Nash,

To answer your questions:

  • EMC uses a pairstyle hybrid to include charge calculations, which combines the standard dpd pair style with a charge model I developed under dpd/charge as described in Weiss et al., Annu. Rev. Chem. Biomol. Eng. 2016, 7, 65-86. This solution causes the order to appear as you described.
  • Due to the coarse-grained nature of DPD, charges are typically variable and not constants associated with specific types as the ITEM MASSES section describes. As a consequence, DPD force field files themselves do not contain charges. What you describe as a ‘workaround’ is actually the intended way of defining charges when using force fields like DPD, i.e. charges are defined by SMILES as provided in the ITEM GROUPS section of your .esh.