[gmx-users] Simulation Techniques***Energy minimisation for Small Molecules.

[Mark Abraham]

jaimohan sm wrote:
> Helo to all Gromax users...
> I am very interesting to working in small molecules...ie calculating 
> energy in the sense by using thermodynamic laws.
> Is there is any chance to minimise the energy for a small fragment of 
> proteins ..may be ligand binding site or an cavity portion.
> Without disturbing the actual geomentry of the protein structure and 
> function.

Your chance of accurately corresponding to real physics depends 
critically on your model physics. Proteins are too large to use good 
physical models, so quantum-mechanical/molecular-mechanical (QM/MM) 
hybrid methods have evolved to treat a component at QM level and the 
bulk of the structure at MM level. This is done because the steric 
constraints and electrostatic field of the rest of the protein affect 
the small region of interest. If you don't care about it, then give up 
and do small-model QM on your region of interest. I understand gromacs 
can do this sort of QM/MM calculation in concert with the QM package 
GAMESS, but know nothing about it. The QM package Gaussian can also do 
it (but not with gromacs), there may be others. You will need to read 
the literature for a good understanding of the limitations of the 
method. This is a much more of an "experts-only" field than general MD. 
In particular, accurate transition structures are hard to locate.

Mark