[gmx-users] how to add potassium ions correctly?

[Mark Abraham]

Hu Zhongqiao wrote:
> Hi,
>  
> I want to add both sodium and potassium to a protein solution. But there 
> is always error information about potassium after I used genion to add 
> K(K+) and set right no of molecules at the bottom of .top file:
>  
> *No such moleculetype K (or K+)*
>  
> no matter I use K or K+ as potassium name and no matter I used g43a1 or 
> OPLS-AA ff.

You need to #include "ions.itp" and make sure the contents define such a 
molecule type, and that your structure file matches the appropriate 
field and that the .atp file matches the other appropriate field.

> By the way, I used v3.3 and selected gromos 43a1 force fields. I have 
> searched the mailing list and found some confusing information. For 
> example, somebody said the 43a1 can recognize potassium, but someone 
> else said it can not and suggested to use OPLS-AA instead. So can 
> anybody who have experiences give an authoritative explanation and 
> solution for using potassium?

This is one purpose that reading the original literature on these force 
fields is good for.

> Further, is there any reference which provides a general description for 
> all force fieds in Gmx v3.3, which is expected to be more detailed 
> than something showing in brackets after the ff names  when using the 
> command pdb2gmx, so that users can know their advantages and 
> disadvantages and select a suitable one according to different aims?

I think that would be a false economy. The strengths and weaknesses of 
these force fields can't be summarised readily for all applications. 
There really is no substitute for doing your own survey of the 
literature with your application in mind.

Mark