[gmx-users] non-specific protein-protein interactions

[Mark Abraham]

On 19/11/2011 6:42 AM, Ben Reynwar wrote:
> On Fri, Nov 18, 2011 at 7:23 AM, Mark Abraham<Mark.Abraham at anu.edu.au>  wrote:
>> On 18/11/2011 8:10 AM, Ben Reynwar wrote:
>>> Dear gromacs list,
>>>
>>> Does anyone know how well current force fields capture the energetics
>>> of non-specific protein-protein interactions?  I'm simulating a
>>> protein with an arm that alternates between (apparently) non-specific
>>> adsorption to the main-body of the protein and free movement in the
>>> solvent.  Can I place any trust in the results of an MD simulation for
>>> something like this?  I haven't seen any comparison of experimental
>>> results with simulation results of the energetics of non-specific
>>> protein-protein interactions so I'm a little skeptical about it.
>>> Currently I'm using generalized born implicit solvent, which is
>>> perhaps a mistake when solvation energies could be critical to the
>>> results.
>>>
>> I'd be very skeptical unless you can observe many adsorption/desorption
>> processes and with explicit solvent (though they're probably mutually
>> exclusive).
>>
>> Mark
> I think I asked my question poorly.
> In a perfect simulation with explicit solvent and infinite sampling,
> roughly what kind of error might one expect for peptide-peptide
> interaction energies due to the force field?
> Presumably there are papers out there that go into this, but I'm not
> finding them.  If anyone knows of any directions to point me in, I'd
> really appreciate it.

I don't recall having ever seen an explicit solvation simulation on even 
a small peptide that claimed exhaustive sampling, never mind a system 
that shows such complex behaviour. It's possible someone has used 
slow-growth methods to grow a small peptide ligand onto the side of a 
protein, and that could give a guide to the accuracy of inter-protein 
energetics, but I've never gone looking for such.

Mark