Re: [AMBER] TI +- variance is coming out to be very high

From: sunita gupta <sunita.bio.gmail.com>
Date: Tue, 5 Oct 2010 14:39:28 +0530

Hello Thomas

I am doing four independent simulation in two different environment:
1. Ligand in water environment
1a. charge annihilation by masking charge on ligand in one group
1b. vdw decoupling by scmask ligand in one group (keeping charges in both
groups to zero)
2. Protein+Ligand in water environment
2a.charge annihilation by masking charge on ligand in one group
2b. vdw decoupling by scmask ligand in one group (keeping charges in both
groups to zero)

I plotted the DV/DL values of all the four steps the charge annihilation
step has converged properly and the vdw as not, kindly find the plots
attached here with this mail.

I am also running few more lambda values for vdw decoupling in order to get
better convergence.










ligand in water (with masked and unmasked state) and protein
I plotted the values of DV/DL in order to see the convergence....The Charge
annihilations step converges very well for both (ligand in water) and
(ligand+protein complex in water)..Vdw decoupling is not showing proper
convergence

On Tue, Oct 5, 2010 at 1:04 PM, <steinbrt.rci.rutgers.edu> wrote:

> Hi,
>
> from the naked values alone its hard to see what may have gone wrong here.
> Did you check your simulations for convergence? What kind of change are
> you simulating? Do plots of your dVdl values come out smooth and suitable
> for numerical integration?
>
> Kind Regards,
>
> Thomas
>
> Dr. Thomas Steinbrecher
> BioMaps Institute
> Rutgers University
> 610 Taylor Rd.
> Piscataway, NJ 08854
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>



-- 
-- 
SUNITA GUPTA



_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber

Received on Tue Oct 05 2010 - 02:30:03 PDT
Custom Search