Having all pairtypes defined in the topology file

User discussions
1

Dear gromacs users,

I am planning to run a simulation of a polymer in water. As the polymer is a non-standard polymer, I generated the CHARMM36 parameters for the molecule from CHARMM-GUI server. Now I will have to a simulation of the polymer in water.

The forcefield.itp file of the polymer contains only the necessary atomtypes, pairtypes etc., for the polymer. So, to prepare the topology of the polymer solvated in water, I need to incorporate tip3p water atomtypes, bondtypes, angletypes, pairtypes etc. I have added the atomtypes, bondtypes and angletypes for the tip3p water model. Now what I am wondering is about the [ pairtypes ] directive. In the forcefield.itp file I have already set gen-pairs = yes;

My first question is, in this case, do I have to manually add the pairtypes entries by searching ffnonbonded.itp file of original or full charmm36 forcefield that has water atoms? For example, water has atomtype: HT and OT; now there are many pairtypes entries present in the charmm36 (full) forcefield including OT and many other atomtypes, HT and many other atomtypes. As I am setting gen-pairs = yes, even if I am not manually including all pairtypes entries with respect to HT and OT, will that be accounted?

My second question is: Is the pairtype relevant if two atoms are actually part of two different molecules? As water atoms and polymer atoms are part of two different molecule, do I have to worry about the pairtypes including water atomtypes and some of the polymer atomtypes?

Any help or suggestion will be much appreciated, thank you.

2

In most force fields, including CHARMM, the parameters pairtypes are only used for 1-4 interactions. Water doesn’t have 1-4 interactions, so pairtypes for water atom types are irrelevant.

3

Thank you for for the explanation Prof. Hess

4

I understand that these pairtypes (in this case water) are irrelevant, but what about if they are included in the .itp file? Will the software ignore them, or will their inclusion cause artifacts in the simulation?

5

Any parameters defines for types which are not used in moleculetypes are irrelevant.