[gmx-users] Explicit Topology Question

[Mark Abraham]

Hi,

I don't think there's a way to make pdb2gmx do that automatically, but a
simple post-processing script should do the job for removing such angles.
IIRC grompp will just eat whatever is now in the topology.

I would definitely invest time in whatever you can do to ensure that your
model descriptions agree, eg that single point energy angle contributions
can be reproduced from whatever originally implemented the model.

Mark

On Mon, 9 Jan 2017 21:24 Thomas Allen <thomas.c.allen at rice.edu> wrote:

> Let me include an example. I have three CG sites, labeled S, R, and C,
> and so a portion of a chain containing three of the monomers would
> appear schematically as:
>
> S1   S2   S3
>   |    |    |
> R1---R2---R3  -> chain continues
>   |    |    |
> C1   C2   C3
>
> The issue in this case is that both the angles R1-R2-C2 and C2-R2-R3
> are defined. The parameter set I have inherited (and don't have a good
> way to change) seems to only refer to the latter type of angle when
> considering C-R-R angle interactions (i.e. the angle is specified with
> reference to the 'forward' direction along the chain, so I only want
> C1-R1-R2 and C2-R2-R3 here, along with the obvious continuation with
> C3 and later monomer units).
>
> What I would like is some way to prevent GROMACS from generating the
> 'backward' angle R1-R2-C2, which it currently does automatically. I
> considered your suggestion regarding the bonds, but the only ones I
> have defined and which show up in the topol.top file after processing
> are those required to hold the structure together as shown above; I
> can't find any redundant bonds in this case.
>
> I realize what I'm asking for is probably somewhat contrary to
> standard usage in all-atom MD, but I was wondering if there was a way
> to support it anyhow.
>
> Thanks,
> Tom
>
> Quoting Mark Abraham <mark.j.abraham at gmail.com>:
>
> > Hi,
> >
> > Do you mean on the same three atoms, or that each of your beads has such
> > "backward" angles? I haven't looked at that code in a long time, but it's
> > probably pretty dumb - two bonds with a common atom generate an angle, so
> > if you don't want the angle, don't specify redundant bonds?
> >
> > Mark
> >
> > On Mon, Jan 9, 2017 at 7:30 PM Thomas Allen <thomas.c.allen at rice.edu>
> wrote:
> >
> >> Hello all,
> >>
> >> I'm currently implementing a simple polymer coarse-grained model in
> >> GROMACS through some modifications to the OPLS force field. An issue
> >> I'm having is that there are certain angles generated when parsing the
> >> topology file which I don't want. For example, assuming atoms A, B, C,
> >> I get the angles:
> >>
> >> A B +C
> >>
> >> and
> >>
> >> -C B A
> >>
> >> However, I only want the forward version (A B +C). I know that
> >> unwanted dihedrals can be excluded by topology file settings, and I
> >> know that CHARMM has settings which allow only certain angles or other
> >> interactions specified explicitly in the topology file entry to be
> >> generated for a given residue. I was wondering if there was an
> >> equivalent when using GROMACS, or if I will have to come up with some
> >> clever text editing to fix the issue in the topol.top file itself.
> >>
> >> Thanks,
> >> Tom Allen
> >>
> >> --
> >> Gromacs Users mailing list
> >>
> >> * Please search the archive at
> >> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> >> posting!
> >>
> >> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >>
> >> * For (un)subscribe requests visit
> >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> >> send a mail to gmx-users-request at gromacs.org.
> >>
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> > or send a mail to gmx-users-request at gromacs.org.
>
>
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>