[gmx-developers] Reaction Filed crash!

[David van der Spoel]

On 12/18/11 12:48 AM, baptista at itqb.unl.pt wrote:
>
>> 3.
>>> We can also discuss if, as a general rule, using a continuum reaction
>>> field is better or worse than using a lattice method such as PME, but
>>> that
>>> is a different issue. Like many other people (Wilfred van Gunsterem,
>>> Arieh
>>> Warshel, Alan Mark, Philippe Hünenberger, etc), I'm really not convinced
>>> that lattice methods offer any real advantage, because I find the
>>> published evidences for either their benefits or their artifacs to be
>>> rather weak or contradictory.
>>
>> For HOMOGENEOUS systems, then I agree; the configurations sampled can be
>> achieved by cheaper methods than PME -- basically, the configurations are
>> dominated by short-range effects, and beyond a certain range, all that
>> the
>> extra interactions do is affect the overall energy, which can be
>> calculated
>> pretty well by many methods (including reaction field).
>>
>> When lattice methods are needed is inhomogeneous systems, such as
>> membranes
>> and interfaces, because the long range order affects the configurations
>> sampled, and RF and other continuum methods simply can't handle those
>> well;
>> properties will depend on cutoffs, etc.
>
> Indeed, the symmetry of those systems is totally different and much more
> in line with a lattice model than with a a spherical isotropic RF.
> Still, the artificial periodicity along the normal to the interface may
> sometimes lead to strange effects, as in the case of a charged bilayer
> (charged lipids and/or adsorbed counterions). We know that the electric
> field should not decay as you move away from one of the monolayers
> (charged plate), because the oblique contributions from the infinite
> surface exactly cancel that decay -- a somewhat counter-intuitive effect
> due to the long-ranged order effects that you mention, and which would
> never be captured by a standard RF. However, in addition to
> realistically mimicking the infinite interface, a lattice model also
> places in the normal-oriented boxes a parallel copy of the oposite
> similarly charged monolayer, so that the field in the solvent region
> would cancel (two identically charged parallel plates), and so we lose
> the constant field that is so typical of a charged interface... So,
> lattice models seem a mixed blessing in this case. Actually, we are
> currently considering and testing alternative approaches to use with
> charged membranes, so any thoughts are welcome.

If you have a charged membrane with sufficient counter ions (and salt) 
you have a dipolar system and the field will fall off with distance from 
the membrane, if you leave out the counter-ions it will remain constant. 
Using PBC the field will be enhanced artificially. We have derived an 
analytical relation for this effect in the supporting information of 
this paper: http://dx.doi.org/10.1002/anie.200703987

Your best bet is to establish using PME and counter-ions how large the 
PBC effect is, then you can probably change the epsilon_surface slightly 
to reduce the effect. The exact value should depend on the box size. I 
have not tried this myself though.

>
> Best,
> Antonio
>
>
>
>


-- 
David.
________________________________________________________________________
David van der Spoel, PhD, Professor of Biology
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++