optimizing parameters for a complex and its pieces

[eft rsd]

My question is rather general and not exclusive to CP2K, however, and as I 
am new to simulations, I hope to get advice from people with experience in 
both simulations and CP2K. 

Suppose I have two species, A & B, and that I want to calculate their 
binding energy upon making the complex AB. This value is given as (total 
energy(AB)-[total energy(A)+total energy(B)]).

Do we need to 

1-optimize the MGRID parameters for A, B and AB separately? 

2-or is it enough to use the optimized parameters for AB and apply them to 
A, and B? 

What if I want to compare the structure of A and B before and after 
complexation? Would using different sets of optimized parameters for 
individual A  and B and then for AB still allow reliable comparison of the 
structural parameters? 
How good/bad would that affect the comparison? 
  
Thanks,
ER
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20151127/0c3b33b7/attachment.htm>