Hi,
I am trying to implement some glucose residuals using GROMOS56ACarbo forcefield (Hanssen et al. 2010, * DOI: 10.1002/jcc.21675), and I ran into the problem of needing to implement separate LJ interactions between specific pairs of atoms that override the interaction parameters of their corresponding atomtypes. Take an example:
In a glucose monomer, the hydrogen atom on O1 (name HO1, type H) and the oxygen atom O5 (name O5, type OR) have specific interaction parameters (see entry N1, Table 5 on Page 1008 of the cited paper). For two such atoms in the same monomer (same “residue” as I will implement in the .rtp file), they should not follow the interaction parameters between type OR and type H, but instead use C6 = 0.0 and C12 = 0.7e-6. How can I make this happen? I obviously need to add this pair of the atoms to the [exclusions] directive in my rtp file. But what else should I do so that GROMACS won’t ignore this pair, but apply a different set of interaction parameters from those deduced from the atomtypes?
Thanks. A follow-up question, does rtp file allow a [pair] section so that a pair of atoms in the residue can be populated to all the resiudes in the topology file? (I don’t think rtp file allows such directive) If not, what is the easiest way for me to specify an arbitrary but consistent pair of atoms in every residue in my simulation (say the “O5-H1” pair occuring in every monomer), without manually adding them in the topology file myself? Asking because our current workflow relies heavily on pdb2gmx, so we would like to avoid postprocessing the topology file ourselves again if possible.
I now realized that I don’t understand what you are asking. I thought you were asking about 1-4 interactions.
If you are talking about 1-4 interactions, add the pair of types and LJ parameters in the [ pairtypes ] section in the force field (not rtp) file.
If you are talking about non-bonded interactions, add the pair of types and LJ parameters in the [ nonbond_params ] section in the force field fil.e
In either case you only need to add the pair of types and parameters once, that will set the parameters for all interactions between any pair of atoms of these types.
Hi, I found this post while searching for a doubt of mine. It seems you answered it, but I would like to confirm.
If I want to implement a specific interaction for two given atom types (a pair) that is not the one arising from the combination rule for their LJ parameters, I should use the [ nonbond_params ] section, right? The [ pairtypes ] section is for the 1-4 interactions only? That is a litle confusing. So, I want the two hypothetical atom types, lets say A and B, to interact with all other atom types via combination rules of their own LJ parameters, but I want their specific A-B interaction to be given by diferent LJ parameters (like for a stronger, quasi-chemical interaction, for example).