[gmx-users] Parametrisation of the cyclic nucleotides in Gromos force fields

[Justin Lemkul]

On 12/16/12 12:52 AM, James Starlight wrote:
> Justin,
>
> It's not quite understood for me why such errors occurs in the atoms
> of standard residues when I've bounded them to the C term of my
> chromophore if the geometry of the adjacent residues might not be
> changed. So it likely that some errors occur during parametrization of

Ignore geometry.  Focus on atom types.  Bonded interactions are assigned based 
on what atom types are used for the atoms involved.  Their names, coordinates, 
and all that other stuff are irrelevant for the purposes of pdb2gmx.

This email thread has gone on for dozens of messages now, all with basically the 
same problem - so you need to isolate your focus.  It doesn't matter if you're 
getting two errors or a hundred; work one-by-one to identify each offending 
interaction type and determine what might be causing it.  There's nothing 
"standard" about what you're trying to do, and there is no guarantee that there 
are bonded parameters built-in for the types of interactions you're looking for. 
  In theory, the topology obtained from Swiss-Param should have everything you 
need in one form or another, such that parameters can be added in, if needed.

> the molecule which I could not identify ( In any case I'll be very
> thankful for anyone who can provideme with some other version of the
> integrated chromophore in charm for my future effort-on-errors :)

Surely a search of the literature would turn up people who have simulated GFP 
using Gromacs, no?  Contacting their corresponding author(s) is probably a very 
efficient means to establishing what others have done.

> The main problems which I've forced in the improper group definition (
> I've not recognize in the paper of simulation in charmm22 what groups
> should define as improper although Swiss param produce the big set of
> such groups- see below )
>

Any planar group (peptide bond, aromatic ring, etc) should be assigned an 
improper.  These are described in the manual.

> In addition It'll be very hardly to connect chromophore with adjacent
> residues mainly due to definition of the addition dihedrals.
>
> Finally Im not quite sure about CMAP definition ( I've used
> C NH1 CT1 C NH1 1 24 24\
> C NH1 CT1 C N 1 24 24\
> in the CMAP.itp with the atom names of chromophore
>
> C NH1 CT1 C NC=O 1 24 24\
> C NC=O CR C=O N=C 1 24 24\
> C=C C=O NC=O CR C=O 1 24 24\
> C=O NH1 CT1 C NH1 1 24 24\
>
> where NC=O and N=C types correspond to  the N and CR is the C-alpha atoms

Again, names are irrelevant.  Only the types matter.  Note that each CMAP entry 
corresponds to a huge sequence of numbers (note that '\' is a line continuation 
character), so I'm not sure what you're trying to do in the cmap.itp file.  If 
deriving CMAP parameters becomes prohibitive, you may need to consider a 
different force field, though all of the atoms involved should be in the backbone.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================