[gmx-users] umbrella sampling with position restraint

[chris.neale at utoronto.ca]

Dear Víctor:

what you should do depends on what you want to get. I suggest that you  
simply cut your PMF at the point where the bilayer is "lost" and  
explain why it is cut. If you really want to pursue your idea, then  
you need to have a good grasp of exactly what you are doing by such a  
method. You would want to start by calculating the free energy of  
applying this additional restraint to a bilayer that is not  
interacting with your solute. The problem then is that you would  
really need to also calculate the free energy of releasing this  
restraint upon reaching the center of the bilayer, and it sounds like  
that not only unfeasible, but also uninstructive. My last thought is  
that you'll want to ensure that the destruction of the bilayer is  
actually the equilibrium action of your solute. Perhaps you simply  
don't have the equilibrated orientation, etc, that would allow this  
system to be stable. In that case, you might consider using your  
method as a way to generate the starting structure, which would then  
be simulated in the absence of this additional restraint for data  
collection.

In the end, you need to have a clear idea of what you want the PMF for  
and also have a clear idea of your path fro the unrestrained state to  
your umbrellas. Free energy is a state property, but you need to  
understand the full path, and if you forget to include the cost of  
something like adding the additional restraint to a neat bilayer, then  
your study will be of little value. Note that the relative free  
energies in the state with the additional restraints are not useful in  
and of themselves just as the intermediates of any alchemical  
transformation are not useful on their own and are only useful as a  
means to get to the endpoint.

Chris.


-- original message --

I am using umbrella sampling method for calculate the PMF for one
molecule in a bilayer system made of SDS and decanol. when the molecule
is in the middle of the bilayer, generate a hole and the bilayer system
is lost.

I am thinking use position restraint for one of the atoms in the SDS  
(sodium dodecyl sulfate). is this a good idea?

any suggestion will be well received



Thanks in advance

Víctor E. Bahamonde Padilla
Laboratorio de Fisicoquímica Molecular
Departamento de Química
Facultad de Ciencias
Universidad de Chile
Phone: 562-978-7443
veduardo at ug.uchile.cl