Re: Free energy calculation under external bias, (e)ABF/US/etc

From: Jérôme Hénin (jerome.henin_at_ibpc.fr)
Date: Fri Jul 28 2017 - 12:17:05 CDT

On 22 July 2017 at 12:29, Jim <jim.jim.strong_at_gmail.com> wrote:

> Hello,
>
> Thanks for your comments - very helpful.
>
> I have another question about additional biases when performing ABF of a
> small molecule, let's say for simplicity ion, permeation through a channel.
> If the system has a number of other small molecules (e.g. other ions, small
> molecules, etc) that can enter the channel and interfere with the sampling
> of the permeating ion within ABF windows in the channel, the PMF can
> drastically change and become inaccurate. In that scenario, I assume it's
> recommended to keep all other molecules (except water) out of the channel
> during the ABF sampling runs. Is this a common practice?
>

There is no accepted practice there, because that is a modeling choice that
will affect what process you're actually studying. If you decide other ions
in the channel are not that important, feel free to exclude them - then be
ready to explain you decision to other scientists. That is a scientific
choice, not a technical one (of course it has technical consequences).

If so, what is the best way to do that? Should harmonic constraints be
> applied to each molecule separately in the NAMD config file, or some
> colvars be used instead, or tcl forces? Can one put "wall" boundaries at
> the top and bottom of the protein channel as to disable entrance of
> specific molecules? Is there a better or commonly adopted way of how to do
> that? Obviously, ease of use combined with negligible slow down of the ABF
> simulations would be the preferred way.
>

Depending on the number of solutes involved, that would be either one
colvar per solute, or possibly a coordNum colvar, or a tclBC script if
scaling becomes an issue.

Jerome

This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:30 CST