Re: [AMBER] Umbrella sampling coordinate question

From: Carlos Simmerling <>
Date: Wed, 9 Mar 2011 07:04:24 -0500

you haven't made it clear what you're trying to calculate. Are you following
a procedure shown to work well in a peer-reviewed article? I suspect that
what you want to do is quite challenging. Keep in mind that umbrella
sampling works well ONLY if the reaction coordinate you choose is the one
responsible for the relevant energy changes, and you must be able to fully
sample over motions in all of the other degrees of freedom in each of the
umbrella windows. I suspect that at a fixed distance you will find it
difficult to properly sample over all other motions (such as side chains,
ligand flexibility, etc.

given that, I'm not sure quite what you're asking. It seems that you have
different numbers of waters. you can adjust the water buffer length in leap
when you build the molecule- it might require trial and error.

during your umbrella runs, make sure that you have overlap in the
distribution of distances (histogram them) between all neighboring windows.
any gaps will result in errors in free energy profile.

On Tue, Mar 8, 2011 at 6:21 PM, Rajesh Raju

> Dear Amber users,
> I am planning to do an umbrella sampling on one of the receptor
> molecule. I used the distance between the center of masses of the
> receptor and ligand molecule as constraint. My doubts are:
> [1] Do I need to make different coordinate files for the complex,
> corresponding to the distance between the center of masses of the two
> molecules? or Can I use the same coordinate file for all individual
> simulations of the different windows?
> [2] changing the distance variable in the 'DISANG=dist.dat' file to
> different distance values corresponding to different windows, and keep
> the same coordinate files with center of masses of two molecules close
> to zero, would be the same effect as that of the doing different
> window simulations, making new coord files corresponding to the
> umbrella sampling distance.
> So far I proceeded with the second approach:
> I made the new coordinates corresponding to different R variables from
> 0.5 to 10 ..20 windows. and made parameter and coordinate files for
> the individual windows..The number of water molecules are slightly
> different....
> Which is the best way?
> Thanks
> Rajesh
> _______________________________________________
> AMBER mailing list
AMBER mailing list
Received on Wed Mar 09 2011 - 04:30:04 PST
Custom Search