[gmx-users] tpr without top-file

[Daan van Aalten]

Hi Peter & List

We are currently working on producing all atom topologies with PRODRG,
initially for the gromos96 ff and later for OPLS and others.
This involves some major surgery deep inside the PRODRG brain, but I am
expect we will have some initial supports for these force fields some time
around the end of november.

However - this of course has nothing to do with "two interacting
molecules" - topologies are for molecules only and not whole systems - you
simply need to have a topology for each of your individual molecules and
then build a system topology from that.

Daan

On Fri, 1 Oct 2004, Peter Zoon wrote:

> The problem with prodrg is that it removes all my hydrogen atoms and that
> it will not accept mocules of more than 2 parts. A rotaxane is a molecule
> which consists of a linear part, mostly saturated hydorcarbon (the
> thread) and a ring which goes around the thread. This ring (aka
> macorcycle) is held in place at one position by hydrogen bonds the tow
> ends of the thread are then blocked by bulky groups  and the ring is
> mechanically locked onto the thread. (see Brouwer et al SCIENCE 291
> (5511): 2124-2128 MAR 16 2001). By applying a photochemical stimulus the
> macrocycle moves from one position to another (since the hydrogen bond
> affinity is changed) and back again.
>
> So i need all the polar hydrogen atoms and prodrg deletes them. Writing a
> topology by hand is I think a rather time consuming job for a system of
> more then 250 atoms. This besides the fact that i would'nt really know
> how to write a topology from scratch.  Or could i take the topology from
> prodrg and easily add some hydrogen atoms?
>
> Peter
>
> David van der Spoel wrote:
>
>  On Fri, 2004-10-01 at 12:07, Peter Zoon wrote:
>
>
>  Hi all,
>
> Here's the story. We wrote a program which converts trajectories from
> Macromodel to xtc. We obtain a initial gro file and the trajectory in
> xtc format. We would like to use the gromacs analysis routines to
> analyze the trajectory. We a looking at a rotaxane in particular we want
> to study the hydrogen bonds between the macrocycle and the thread and
> the dihedrals within the thread.
>
>
>  the problem I presume it the topology. How about generating one using
> prodrg? Many analysis tools can substitute the tpr file for a gro/pdb
> file but g_hbond needs the atom names and the connectivity.
> g_dih actually also needs the topology, although some things can be done
> with just an index file (g_angle, which also analyses dihedrals)
>
>
>
>