[gmx-users] Charges for Coulomb potential

[Mark Abraham]

On 30/08/2012 3:18 AM, Delmotte, Antoine wrote:
> Many thanks for your quick response, Mark.
>
> So epsilon_r is actually different from the actual relative 
> permittivity (which would be like 80 for water), is that right?

You can have an atomistic solvent or a continuum solvent, but not both 
at once.

>
> If so, how does one usually chooses this value of epsilon_r? Is the 
> default value of 1 applicable for most cases (like for simulations in 
> water), or should one obtain it from the literature somehow?

The force field defines it (normally to 1).

Mark

>
> Thank you,
>
> Antoine
>
> On 29/08/12 14:47, Mark Abraham wrote:
>> On 08/29/2012 11:08 PM, Delmotte, Antoine wrote:
>>> Dear Gromacs users,
>>>
>>> I would like to know which charges are used by Gromacs in the 
>>> calculation of electrostatic interactions in the standard coulomb 
>>> potential:
>>>
>>> E_electrostatics = 138.935 * q1 * q2 / (epsilon_r  * r_12);
>>>
>>> Are q1 and q2 the values tabulated in the file "aminoacids.rtp" of 
>>> the force field used?
>>
>> If you use pdb2gmx to generate your .top, yes. More generally, 
>> GROMACS uses the charges found in the .top.
>>
>>> Could you also please redirect me towards the relevant literature 
>>> for the calculation of epsilon_r in Gromacs?  (or does Gromacs uses 
>>> a fixed epsilon_r?)
>>
>> You choose it. See parts of manual 4.1 and 7.3
>>
>> Mark
>