CHARMM Development Project Forums User Discussion & Questions Energy terms, Constraints, Restraints, and Solvation rdie vs cdie

Dear all,
In nbonds.doc, their definitions are:
CDIE - Constant dielectric. Energy is proportional to 1/R.
RDIE - Distance dielectric. Energy is proportional to 1/(R-squared)

And they are used in the following minimization examples:
1) minimize sd nstep 50 cdie eps 1.0 fshift vshift cutnb 13.0 ctofnb 12.0 !Copy from professor Lennart Nilsson's minimize.inp(scripts forum)
2) minimize conj nstep 500 tolenr 0.001 cutnb 14.0 rdie eps 2.0 vswitch shift

My questions are:
a. If I use "RDIE", then what is the dielectric constant of all the selected atoms? What does the "distance dielectric" explicitly mean? Any formula?
b. If I want to minimize a protein in solvent(water), what should I use? And How?

Thank you very much!

When thes system is solvated in explicit water, cdiel and eps=1 is appropriate. Distance-dependence dielectrica (rdiel) is a way to screen out charges and for systems in vacuo (mainly for mimizations).

I am using implicit solvation. Could you specify how rdiel screening out charges? It is claimed somewhere in the documents, but I don't know how.
And if I want to minimize the protein in an implicit solvation, which one is better and how to choose? Any references or comments will be welcome.

Minimization with the solute in an implicit media would (as far as I understand) only require that you specify the dielectric constant (i.e not equal to 1). I don't think a rdiel option would be stringent in this case (or in any cases at all actually).

Thanks!

The dielectric constant is a continuum concept and one cannot really speak of the dielectric constant of an atom.
To answer your question about RDIE: The standard way to evaluate electrostatic interactions in CHARMM is to use Coulomb's law for all nonbonded atom pairs
E= qi*qj/(4*PI*epsilon0*epsilon*r)
With RDIE this take the form (interatomic distance r is measured in Angstroms)
E= qi*qj/(4*PI*epsilon0*epsilon*r*r)
the value of epsilon (default 1.0) can be changed with keyword EPS <real>.

Thank you very much as always!

It is noteworthy that the electrostatic "energy" with the rdiel-formulation actually is the expression for the electrostatic (Coulombic) force...

Dear Prof. Nilsson,

I have a question on RDIE.
What is the best value for eps (default=1.0) when RDIE is used in vacuum to mimic solvent effect ?
Is this depend on a system ? If then, how can I choose the eps value ?
Thank you in advance.

I have not really used rdie to mimic solvent in very many years, and so cannot advice about a value for EPS; your best choice is probably to use a more advanced implicit solvent model, of which there are several in CHARMM.