Re: [AMBER] unique atoms in TI softcore

From: Hannes Loeffler <>
Date: Thu, 17 Oct 2013 15:42:58 +0100

On Thu, 17 Oct 2013 07:45:04 -0400
David Case <> wrote:

> > On Oct 17, 2013, at 6:01 AM, Hannes Loeffler
> > <> wrote:
> >
> >
> > I am currently trying to find out how to determine the unique atoms
> > for the TI softcore approach.
> The soft core mask input variable is used to specify the unique atoms.
> ...not sure if this helps, but it seems to be what you are looking
> for...dac

Not quite I'm afraid because I still don't know what you mean by
"unique". I tried to explain my thoughts in my previous email but
maybe I was not clear.

Maybe you mean scmask should contain only those atoms which are absent
in the respective other parm file and thus could also be just defined
explicitly as dummy atoms (as for the second TI implementation that is
supported in Amber12). But why does the tutorial make H and O softcore
atoms in the benzene to phenol transformation where H obviously becomes
O: is that because the elements or atom types are different, or because
they belong to a "group"?

We are currently writing an automatic setup tool for this type of free
energy calculation, probably very similar to FEW. If I understand their
paper correctly they leave the decision which atoms should be softcore
to the user. I want to explore if there is a (sensible) automatic
procedure to do this or if it should be left open.

Many thanks,

Scanned by iCritical.
AMBER mailing list
Received on Thu Oct 17 2013 - 08:00:02 PDT
Custom Search