[gmx-users] pair interaction in ligand topology file
dmobley at gmail.com
Wed May 31 16:24:54 CEST 2006
> Yes i mean [pairs] section. I have read in manual this include 1-4 interaction, while using Oplsaa forcefield you need not to specify pairs as they are automatically generated and if you are using RB dihedral this section should be empty. I have few more question. The question may look silly but i m not able to get them fully.
As I said, the pairs list is used to *modify* the usual 1-4
interactions. That is, every force field has a 'combination rule'
which is used to calculate van der waals interactions between atoms.
This usually is represented by a number (1, 2, or 3) as described in
the manual. Entries in the pairs list override these default
interactions. The reason for doing this is that many FF were
parameterized with modified 1-4 interactions, hence the 1-4
interactions need to be modified to work appropriately.
> 1.How it automatically generate pairs? Is it just the set of all 1-4 LJ pairs? Are, hydrogens also included?
All LJ interactions between pairs of atoms are generated using the
combination rule based on the LJ parameters for each atom type, as
specified by the force field or your topology. You should see the
section on combination rules for this.
> 2. What if the cases where 1-5or higher interaction are important. When a pair of atom are in close proximity have more than 3 bonds away. The cases when we have charge species and we r using partial charge for each atom and the total charge is distributed over few distance away
Read the relevant sections in the manual. All of the interactions are
included; the pairs list is typically used only to modify the 1-4
> 3. I m getting positive potential energy without pairs for one of my topology. but with inclusion of some specific non-bonded interacting pair lower the energy to -ve side.
No idea what you're talking about here, but there have been some
responses on the list lately (in response to your e-mails? not sure)
related to positive potential energies. Suggest you see those.
> with thanks
> >Do you mean, "use pair interactions"? If you're referring to the [
> >pairs ] section of your topology, well, you will probably need to look
> >at details on the force field to see what it expects, as the answer to
> >this is force field dependent, and I don't typically use OPLS-AA. The
> >[ pairs ] section is usually used to modify 1-4 interactions, but can
> >also be used to modify other interactions if you do things right. This
> >is explained in the manual.
> >In terms of generating ligand topologies, there are some programs
> >which can do this automatically for OPLS, I think. Schrodinger has
> >one, I believe. Unless you are very well versed in the force field it
> >might be worthwhile looking into this, as obviously, if you make any
> >mistakes in generating your own topology, your results can be garbage.
> > If you are not committed to the OPLS force field, it is also
> >relatively straightforward to automatically generate ligand topologies
> >using AMBER's GAFF force field. Or prodrg can generate topologies for
> >some other force fields.
> >If you are intent on doing it by hand, be sure you know what you are
> >doing, and be very careful.
More information about the gromacs.org_gmx-users