[gmx-users] co-solvent simulation

[Mark Abraham]

On 12/02/2012 4:16 PM, Yao Yao wrote:
> I am simulating a protein in glycerol solution in amber03 ff and 
> followed "http://www.gromacs.org/Documentation/How-tos/Mixed_Solvents".
> Since glycerol is not available in amber03.ff by default, I built it 
> in the ffbonded.itp

You should not need to touch this file to model glycerol. There are no 
atom or interaction types that are materially different from those on 
peptide side chains.

> and aminoacids.rtp.
> For test purpose, I then made a itp for it and solvate it as a protein 
> and changed .top file accordingly, everything seems fine when I grompp,
> but when I insert it as a co-solvent, it shows like that I never did 
> claim any bond angles, dihedrals, .... even when I #include its .itp 
> and change #mols in my new topology.

You need to work out what is different in the two cases. The diff tool 
can be useful here. There is no fundamental difference in the .top 
construction for the two cases you are considering - only in 
construction of the initial coordinates.

>
> I guess the problem is if I do not claim the co-solvent as a protein, 
> ff will not go to ffbonded.itp to recognize the new interactions.
> Is there any other file for co-solvent interactions I need to change?

The ffbonded.itp file is #included regardless of what you name your 
moleculetypes. You may not be #including the correct one - but then 
again I think you should only have the standard one in the first place.

Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20120212/6292faf4/attachment.html>