Re: How to replace NAMD electrostatics

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Thu Nov 21 2013 - 12:02:42 CST

On Thu, Nov 21, 2013 at 6:41 PM, Juba, Derek <derek.juba_at_nist.gov> wrote:
> Thanks for the tips- I've looked into OpenMM and it seems like it actually may be the best platform for what I'm trying to do.
>
> I do have one more question about NAMD, in case I decide to go that way. The electrostatic calculation I'm doing is producing a global electric field, computed from all atoms. But, as you mentioned, NAMD sometimes expects bonded atoms to be excluded from non-bonded force calculations. If I set the "exclude" parameter to "none", it looks like NAMD will think it should be computing non-bonded interactions between all pairs. Is just doing this sufficient to give correct results for non-bonded interactions being computed between all pairs, or would I also have to modify the parameters for the bonded interactions to account for the fact that the non-bonded forces between even the bonded atoms are being explicitly added in?

setting non-bonded to "none" would be wrong. you *do* want those
interactions to be excluded since the bonded interactions are not
changed.

in short, if you do have bonds, you *do* want those exclusions and the
real challenge is going to be to include them into your code (and
doing it efficiently so) and if your system has no bonds, it doesn't
matter.

axel.

>
> -Derek Juba

-- 
Dr. Axel Kohlmeyer  akohlmey_at_gmail.com  http://goo.gl/1wk0
International Centre for Theoretical Physics, Trieste. Italy.

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:56 CST