[gmx-users] Re: gmx-users Digest, Vol 114, Issue 91

[Knapp Bernhard]

> On 10/30/13 11:26 AM, Knapp Bernhard wrote:
>> Hi users,
>> 
>> I was wondering if there is an easy and quick way to get a frame-wise 
>> binding
>> energy between 2 groups of chains (chain ABC vs DE) in an existing 
>> trajectory.
>> This is not indented to be an actual binding free energy approach but 
>> rather a
>> rough indicator if the binding between two chains increases or 
>> decreases during
>> simulation time. This should work like a protein/protein docking 
>> scoring
>> function which assigns binding scores to single conformations. Is 
>> there a direct
>> approach on an existing xtc (even if water was removed and not all 
>> steps were
>> written)? Maybe like g_hbond but in addition with terms like Gvdw, 
>> Gdeformation,
>> Ghydrophobic etc. Alternatively (and probably much more time and 
>> resource
>> consuming), would it be possible to obtain this by the rerun option of 
>> mdrun
>> (adjusted energy monitoring groups (the original setup was "energygrps 
>> =
>> protein    SOL"))?
>> 
> 
> Using energygrps is the closest you'll get without modifying the code.  
> You'll
> get decomposed short-range nonbonded terms, which can perhaps provide 
> some
> indicator of the strength of the interaction between the two entities.
> 
> -Justin
> 

Hmm ok if Gromacs does not provide something like this then I ask the 
question a little bit different: Does someone know a pure scoring 
function for protein/protein interaction? It should not be a complete 
docking program but rather something which can simply evaluate given 
conformations (coming from an MD trajectory) ... and it should also not 
be a webserver (like ClusPro) but a downloadable binary because I want 
to test a quite high amount of conformations. Something like XSCORE [J 
Comp Aid Des, 16: 11-26. 2002] but instead of ligand/protein for 
protein/protein interactions.

Cheers,
Bernhard