Dr Steinbrecher,
The ultimate goal in this project is to mutate the charged aminoacids in
wild type protein to glycine or alanine and carry on the TI to find
relative free energy differences.
When I was looking through literature, there were very few papers that talk
about mutating a charged residue to neutral residue.
One of the papers that I have come across is title "Accurate Estimates of
Free Energy Changes in Charge Mutations", JCTC, 6, 1884.
The authors suggest that when charged species are mutated, due to the Ewald
treatment of electrostatics, there is a self-energy term that appears and
one has to account for it and they suggested a method to do dual mutations.
In their work they have shown it work for Glu to Gly or Ala and mutation of
K+ and Cl- to neutral dummy atoms.
I'm trying to follow that protocol for my charged mutations. When they did
for KCl in water to neutral atoms, they were suggesting that the charged
species in a concurrent dual mutation should be kept at a certain distance.
I quote from the paper.... "In order to guarantee the independence of the
two concurrent mutations, it is necessary to position the counterion at an
adequate distance from the protein in order to minimize the electrostatic
interaction between the protein and the ion. The Bjerrum length is defined
as the distance at which the Coulomb interaction between two monovalent
ions in a uniform dielectric is equal to the thermal energy, kBT".
This was reason for me to set the position restraints for K+ and Cl-. In
the initial file they are separated by 15 A in my case.
On Thu, Jan 19, 2012 at 4:49 AM, <steinbrt.rci.rutgers.edu> wrote:
> Hi,
>
> > When I observed the dV/dL averages for 500ps for lambda = 0.99, the
> > restraint energy is -177.4749 kcal/mol which is the almost all the dV/dL
> > for my system.
> > The reason I had the restraints was to make sure that the K+ and Cl- dont
> > come close during the dual mutation.
> >
> > Is there a better way to approach this?
>
> I am not sure what you ultimately want to do, so it is difficult to
> distinguish better and worse ways to do it here. Decoupling a restrained
> molecule from its surroundings is a different thing from decoupling a
> non-restrained one, with resulting differences in free energy. Both ways
> make sense in terms of different calculations.
>
> If you want the two ions to be restrained in V0 only, then your setup is
> good to go, but shows (unavoidable?) convergence problems at high lambdas.
> If you want the restraints to not count for dvdl, set dvdl_norest, or
>
I want the restraints for both V0 and V1 but since V1 was only with water,
I was not sure how to apply the position restraints for the corresponding
atoms in V1 state.
Is this assumption valid?
I think your suggestion of setting dvdl_norest can be helpful in my case.
I'll set up some calculations and report here.
Thanks for your help.
Regards
Sai
> define the restraint as a distance between K and Cl instead of cartesian
> ones. Or, remove the restraints altogether (which, indeed may also give
> you poor sampling as K and Cl can stick together).
>
> None of these things is wrong per se, they just answer different questions.
>
> Kind Regards,
>
> Thomas
>
> Dr. Thomas Steinbrecher
> formerly at the
> BioMaps Institute
> Rutgers University
> 610 Taylor Rd.
> Piscataway, NJ 08854
>
> _______________________________________________
> 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 Jan 19 2012 - 11:30:04 PST
