[gmx-users] Re: CAN'T CALCULATE THE ENERGY OF SYSTEM ONLY

[Mark Abraham]

Please leave GROMACS correspondence on the mailing list where others can 
contribute to it and learn from it. Please do not use all capitals in 
your subject or email - it's the internet equivalent of shouting.

radhika jaswal wrote:
> Respected Sir,
> As u suggested I did the same, but I was unable to calculate it. I used 
> make_ndx in which I selected protein and then I used with grompp with -n 
> option but the energy it is showing is of the whole system not of the 
> protein only.

You need to define energy groups before you run mdrun (or rerun mdrun). 
Only then will it produce such components of the total energy. Look such 
groups up in the manual.

> To mke it clear suppose if I have a protein A and solvent system B. I 
> want to simulate the system A + B but I want the energy of A separately 
> with the co-ordinates of A as in A + B and then energy of B with its 
> original co-ordinates in A + B as I want to compare the different 

Extracting coordinates of such groups is easier. The predefined groups 
are likely to be fine for you. Look at the documentation for trjconv.

> conformational states of a protein because editconf or genbox adds 
> unequal number of water molecules. OR I have a method to add same number 
> of water molecules to every state as water molecules which are not 
> interacting also contribute towards energy parameter.

I don't understand why you're seeming to add water molecules to states 
in which you are not interested in water molecules.

Mark