[gmx-users] NOE restraints and undefined hydrogen atoms, virtual interaction sites

[Mark Abraham]

Louic Vermeer wrote:
> On Wednesday 30 January 2008 23:56, Mark Abraham wrote:
> 
>> There's no a priori reason why a mixture of a united-atom detergent
>> sodium dodecyl sulfate and an all-atom peptide *couldn't* work. Neither
>> is there any reason to suppose they *would* work without evidence that
>> the SDS parameters were developed with this purpose in mind and suitably
>> validated. For that, you'll have to read the paper referred to on the
>> page you footnoted [1], I guess.
> 
> Obviously, I read the paper that I referred to in my question. The authors 
> have simulated SDS in water at different concentrations, using a united atom 
> "gromacs" forcefield (they don't mention the forcefield more precisely in the 
> paper or in the topology file itself). They do not mention that the SDS was 
> designed for other reasons than simulations of SDS itself. The validation of 
> the SDS topology file was done by the authors of the paper, to me it seems 
> fine. 

If what this paper has done has validated the use of a subset of ffgmx 
(I presume, check the numbers in their topology for yourself against the 
ffgmx .itp files) for SDS-water mixtures, then that's well and good, 
since ffgmx is presumably also validated for protein-water mixtures, and 
so an SDS-protein mixture has a reasonable chance of being valid. Use of 
ffgmx is deprecated, however, though I have never bothered to find out why.

If this paper has developed their own parameters for SDS and validated 
them for a water mixture, then there's no grounds for supposing they'll 
play nicely with any protein force field. They might, but you'll have to 
prove that against some experimental data.

> Several messages on this list have pointed out that mixing forcefields is not 
> a good idea. To use united-atom SDS with an all-atom peptide, it seems to me 
> that I would have to mix two forcefields. I would prefer to use a united atom 
> forcefield for both the SDS and the peptide. 

Fair enough, but mixing two united-atom force fields is, in principle, 
just as bad as mixing a united-atom and an all-atom forcefield. If it 
isn't validated, you've got a random number generator until proven 
otherwise.

> Which leaves open my question 
> about the definition of NOE restraints (on protons) when using a united-atom 
> forcefield.

Mark