[gmx-users] Virtual sites parameters
Justin Lemkul
jalemkul at vt.edu
Tue Jun 18 19:00:02 CEST 2013
On 6/18/13 12:20 PM, Bastien Loubet wrote:
> Dear GROMACS user,
>
> I have been trying to define parameters for virtual sites in CHARMM lipid,
> and I would like to know if there is something simillar to the [ bondtypes ]
> for virtual sites.
> Specifically the way I specify the parameters now is by having lines in the
> topology such as (for a 3fout virtual sites):
>
> [ virtual_sites3 ]
> ; ai aj ak al funct c0 c1 c2
> 40 39 37 51 4 -0.26909 -0.26909 4.5519
>
> But then I have to do that for every virtual sites in the molecule, even if
> the parameters are the same.
> What I would like to be able to do is:
> [ XXXX ]
> ; ai aj ak al funct c0 c1 c2
> H C C C 4 -0.26909 -0.26909 4.5519
>
> and then
> [ virtual_sites3 ]
> ; ai aj ak al funct c0 c1 c2
> 40 39 37 51 4
>
> Where atom 40 is of type H and atom 39, 37 and 51 of type C and XXXX should
> be are directive similar to [ bondtypes ] but for virtual site.
> Does this XXXX directive exist ? I had no luck checking the manual.
>
Not likely, because virtual sites aren't generally built based on types, but
based on specific atoms. The workaround that came to mind is to make use of
#define macros, like how the Gromos96 force fields define bonded terms. You can
simply:
#define param -0.26909 -0.26909 4.5519
Then, for your virtual sites:
[ virtual_sites3 ]
; ai aj ak al funct c0 c1 c2
40 39 37 51 4 param
Not quite as efficient as defining types, but since that's not an option, it
does save some typing.
-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
========================================
More information about the gromacs.org_gmx-users
mailing list