[gmx-users] Ligand Internal energy

[Mark Abraham]

On 6/09/2012 1:02 PM, Marcelo Depolo wrote:
> But if you could simulate the environment just around the ligand, then you
> could calculate H-bond and angles energies in a quantum perspective. The
> limitation is the lack of integration time,
>
> Or am I too wrong?

Sure, any problem you might treat with an MM force field would be better 
if you could afford a QM-MM treatment. My guess would be that most of 
the value would be in a) not using atomic charges that are known to be 
suited for some other environment, and b) having a polarizable charge. 
QM is just an easy (but expensive) way to achieve that. There are 
methods out there based on using QM fragments to do re-parameterization 
of the MM charge on the fly.

> but that can be avoided using a
> frame that corresponds to the average system pose.

Using a single average (or representative) structure still assumes that 
a single static picture is a valid model, and that's a big assumption.

Mark