QM/MM with no MM parameters

Just Got Here

6:53:56 AM PDT - Mon, Jun 27th 2011
Hi all, I'm a NWChem newbie, trying to perform some QM/MM optimization. I have a big protein which hosts a large cofactor at the active site that is not parametrized in the amber force field. Do I have to provide a full MM parametrization of the whole ligand, even if it would be entirely included in the QM part only? In this case, I'd like to know if providing meaningless force-field parameters would be useful or otherwise change the results of the optimization. My plan would be to include these "random" parameters to have the prepare module generate the correct coordinates/topology file. Is this feasible/meaningful in your opinion? Thank you

Guest -

10:09:27 AM PDT - Thu, Jun 30th 2011
If you QM subsystem is NOT covalently bound you only have to provide van der waals parameters and the rest does not matter. In any case you will have to create fragment(s) file for you QM region. If you QM region is linked to MM than more "stuff" (bond, angle, ...) would be required

Just Got Here

12:39:02 AM PDT - Fri, Jul 1st 2011
Thank you a lot for your help. I have only one covalent bond between the QM and the MM subsystem, I guess I have to parametrize correctly the stretch-bend-torsion parameters for that bond and neighbouring atoms only.. correct?

Forum >> NWChem's corner >> General Topics