From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Wed Jul 07 2021 - 09:31:19 CDT
Hello Ishaan, the list should be copied in.
I agree that a full protein flip is very unlikely to occur under normal
equilibrium sampling. However, you are using an enhanced sampling method
precisely with the purpose of *making such rare events occur more
frequently*. When ABF or similar methods are used properly (i.e. along
collective variables that can quickly exchange the supplied external energy
with the rest of the system), you can observe the structural details of the
mechanism and compute the free-energy cost associated with such rare events.
The methodology and technical details are explained in the literature and
in the code documentation, respectively. For example, you could start from
the latter, which contains references to the former:
https://urldefense.com/v3/__https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html*sec:colvarbias_abf__;Iw!!DZ3fjg!uVUdY8aJ-Q4GC2_4-KKTetJC6_m4HYM1-fxUqer2HmI2Z7uiWTx2xlwlFOanL0H6lg$
After reading those, please narrow your questions down to specific details,
preferably technical ones that traditionally find little space in the
scientific literature (and for which a question to the mailing list is most
suitable).
Giacomo
On Tue, Jul 6, 2021 at 8:39 PM Ishaan Roy <royi_at_uchicago.edu> wrote:
> Hi Giacomo,
>
>
>
> Thank you for the response. We have considered what you said and agree
> that if there were no other natural constraints on the molecule then the
> flipping should be expected. However, in our case, the lipids in the
> membrane would prevent the protein from ever flipping in reality, so the
> mechanism does not seem realistic as you put it. It seems extremely
> unnatural. Does this still mean the flipping should occur in the simulation
> to find the free energy? If not, are there any changes you could recommend?
>
>
>
> Thank you,
>
> Ishaan
>
>
>
> *From: *Giacomo Fiorin <giacomo.fiorin_at_gmail.com>
> *Sent: *Wednesday, June 30, 2021 8:53 PM
> *To: *NAMD list <namd-l_at_ks.uiuc.edu>; Ishaan Roy <royi_at_uchicago.edu>
> *Subject: *Re: namd-l: Membrane Protein Flipping in ABF
>
>
>
> Hi Ishaan, ABF or other adaptive biasing methods will try to overcome all
> free energy barriers, including those preventing flipping. So flipping is
> to be expected at some point, the question is whether the mechanism is
> realistic and the associated energy as well. That depends on how big the
> protein is, whether it's atomistic or CG, and a variety of other factors
> that are specific to your project.
>
>
>
> The upper limit and lower limit that you are using are the natural
> boundaries of the tilt colvar: it makes sense to define a PMF in that
> range. You didn't specify whether you also have walls at those locations,
> but even if you did they would have no effect.
>
>
>
> Giacomo
>
>
>
> On Wed, Jun 30, 2021 at 9:45 PM Ishaan Roy <royi_at_uchicago.edu> wrote:
>
> Hi,
>
>
>
> I am running an ABF simulation on a membrane protein to find the binding
> free energy. In one of the systems, when running ABF on the tilt angle
> colvar, the protein flips over almost 180 degrees within the membrane.
> However, in other systems, the protein stays within a smell tilt angle
> range in the membrane when running the ABF on it. The tilt colvar currently
> has an upper and lower limit of 1 and -1. Could this problem be a result of
> the ABF countering interactions that would normally keep the protein in a
> natural orientation? If so, then to what values should the upper and lower
> limit be changed?
>
>
>
> Thanks,
>
> Ishaan
>
>
>
This archive was generated by hypermail 2.1.6 : Fri Dec 31 2021 - 23:17:11 CST