[gmx-users] Defining charges: [atomtypes] vs [atoms]

[Mark Abraham]

Matt Wyczalkowski wrote:
> Looking through the documentation, it seems like the charge of an atom
> may be defined in one of two ways: in the [atomtypes] directive, or in
> the [atoms] directive.  In the past, I've always made them identical,
> but I have seen examples where they differ.
> 
> Why are there two ways to define an atom's charge, and which
> definition takes precedence? 

Having two ways allows for more flexibility. Some force fields might be 
prescriptive about charges for certain atom types, some might not be. 
Accordingly, [atoms] takes precedence.

> Also, during a free energy calculation
> where an atom mutates typeA -> typeB, are the charge values as defined
> in the [atomtypes] section or [atoms] section used?

I don't know, but I expect [atoms].

Mark