[gmx-users] Re: rvdw parameters (and other mdp stuff)

[Erik Lindahl]

Hi,

>
>The important thing here is the heat flow between different parts
>of the system, which is related to the contact surface area of the
>parts. For example, heat exchange will be abundant between water-
>solvated ions and the water. But between a (largish) protein and
>water it might be minimal. The other issue is heat-generation, 
>which tends to be big(ger) for charged groups. So a (large) protein
>will generate relative large amounts of heat, but not be able to
>exchange that heat with the water. In that case, coupling the whole
>system will lead to a correct average temperature of the system 
>(obviously), but a hot protein (and a slightly cooler water).
>

Right. This is important for cutoff-scheme simulations, where the 
cutoffs can
introduce artificial heat in some parts of the system. In principle you 
shouldn't
have the problem if you use PME in combination with a relatively large vdw
cutoff, but in practice it doesn't hurt to have two groups, so most of 
us still use
it to be on the safe side.

Cheers,

Erik