Dear Hannes
Thank you for your reply. I'm sorry, I only got a part of your reply and
don't see the full reply on the archives yet. Yes. I'm using a similar
multisite Calcium model but the number of dummy atoms is different than
Sept Model you provided the link to.
In PEMED I had 2 Ca ions in the restraint step (residue numbers 90 and 91).
What I do is restraint 90 at the end of the restraint step. Then in
the charge step remove the charge from 91 and in the vdW step totally
remove 91. PEMED calculations worked without any issues. Issues arise for
Sander.
In Sander which I'm doing now, I didn't see the need to have two copies
like that so I only have 90.
What I do is, I first calculate the free energy change for removing Ca from
protein/water [dG1] and then calculate free energy change for removing Ca
from a plain water box [dG2]. To calculate the free energy change of
binding Ca to protein I subtract the two numbers [dG1-dG2].
Removing Ca from protein/water has 3 substeps, addition of restraints,
removal of charge and removal of vdW interactions.
Removing Ca from water has 2 substeps, removal of charge and removal of vDW
interactions.
Thank you!
On Thu, Jun 1, 2017 at 2:56 PM, Hannes Loeffler <Hannes.Loeffler.stfc.ac.uk>
wrote:
> On Thu, 1 Jun 2017 14:16:38 -0400
> Sumudu Leelananda <sumudu.leelananda.gmail.com> wrote:
>
> > Dear all
> >
> > I'm doing thermodynamic integration to find the free energy of Ca
> > binding to my protein. The simulations run without issues when I use
> > pmemd. I use a multisite Ca ion model.
>
> Is this something like this?
> http://septlab.engin.umich.edu/multisite-ions.html
>
>
> > Previously for pmemd timask1 =90 and timask2 = 91 were used for this
> > step (two copies of the multisite Ca ion).
>
> This does not look like a proper mask to me. What are those numbers
> referring to and what is the end state? When you say binding free
> energy (and restraints) I would expect that you would want to carry out
> an "absolute" (standard) free energy calculation in which case your end
> state is "nothing" i.e. you want to dec
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jun 01 2017 - 12:30:02 PDT
