[AMBER] QM/MM Implementation Details

From: Jim William Snyder Jr <jsnyder3.stanford.edu>
Date: Thu, 3 Dec 2015 18:38:35 +0000


I'm working on doing some things with QM/MM and Terachem, and I had a few questions about the details of the QM/MM implementation in Amber.

1) I'm trying to understand how the van der Waals interactions are treated when link atoms are employed. In the manual, it says "The Lennard-Jones interactions between QM-MM atom pairs are calculated in the same way as described in the section above with exclusion of 1-2 and 1-3 interactions and scaling of 1-4 interactions." This led me to believe that the forces are modified in some way from the basic MM force field. However, the manual later states, "VDW interactions are not calculated for link atoms. These are calculated between all real QM atoms and all MM atoms, including the MM link pair atoms." It makes sense that VDW forces are not computed for the link atoms, but the statement that VDW forces are computed for all real QM/MM pairs seems to contradict the statement that 1-2 and 1-3 interactions are ignored. Would someone be able to clarify this point for me?

2) On a related note, VDW forces are never computed between 2 QM atoms correct?

3) I was curious about how link atom electrostatics are handled in Amber. Does amber adjust any of the point charges in the vicinity of the link atoms (e.g. by smearing out the charge with a gaussian)? Also, doe amber do anything more broadly to the point charges at the QM/MM boundary, even when there are no link atoms?

Thank you for the help.


Jim Snyder
AMBER mailing list
Received on Thu Dec 03 2015 - 11:00:04 PST
Custom Search