PC GAMESS/Firefly-related discussion club

Learn how to ask questions correctly

Raul Mera
butil_litio@yahoo.com

I have re-read something about MMQM, and at least in common electrostatic embedding, Only the partial charges from the MM part is included in the QM Hamiltonian, so only electrostatic interactions are treated at QM level. Other interactions are treated at MM level and they have a great importance in geometrical optimizations, but concerning the wave function (or density) of a single point calculation, it would be the same to perform a complete MMQM with electrostatic embedding and perform a QM calculation including only point charges for the "MM" part of the system. Of course, anything more complex than a single point calculation would be different.

Maybe Firefly does something more sophisticated than this standard EE method, and I would be grateful if you could tell me about that.


Cheers,

Raśl


On Thu Feb 12 '09 4:22am, Jim Kress wrote
-----------------------------------------
>The TIP3 solvent WOULD COUNT in the 3500 limit.

>>Given that it is possible to include the point charges (and probably obtain the gradients for this point charges) and considering that the input and output format is similar to that of GAMESS US, would it be possible to use an interface like the implemented in some MM packages for GAMESS US?

>No.  GAMESS US has a totally different input file structure for their implementation of QM/MM and is not compatible with PCGAMESS 6.4.

>Also, I erred in my previous response.  For a single point calculation ALL nonbonded interactions would be incorporated, not just those due to point charges.

>Since the "number of atoms limitations" seems to be an area of significant concern to you, I'd suggest you talk to Alex about how he has overcome this limitation or you look at other QM/MM packages which have no such limitations.