From: Aron Broom (broomsday_at_gmail.com)
Date: Fri Jun 07 2013 - 09:50:12 CDT
I think in this case you don't want to apply the Jacobian correction, but,
it's just a simple mathematical correction that you can do or undo once you
have your final PMF, so it's not a big deal what you choose it to be.
If your PMF appears to have far too deep of a well near it's initial
position, and the simulation seems to not explore that initial region after
all the fullSamples have been filled, you might need to consider rerunning
things with a higher value. At least that is my experience, which is
related to what Jerome mentioned about to much non-equilibrium work being
done when the value of fullSamples is too small.
On Fri, Jun 7, 2013 at 10:43 AM, George Patargias <gpat_at_bioacademy.gr>wrote:
> Hi Jerome
> Many thanks for these comments.
> > Hi George,
> > ----- Original Message -----
> >> Hello
> >> I would like to set up an ABF simulation to study the conformational
> >> change
> >> of a protein using as a colvar the C-alpha RMSD of two conformational
> >> states.
> >> What is the reason of defining the
> >> lowerwallconstant/upperwallconstant
> >> options;
> > Typically with a RMSD, you will only need the upperWallConstant, as the
> > natural tendency is to drift away from the
> >> I am not sure I understand the meaning of the hideJacobian option. Is
> >> it to be switched on or off?
> > In doubt, leave it at its default value (no).
> >> How can I decide on the value of the option fullSamples? By checking
> >> the
> >> PMF convergence?
> > FullSamples affects the early stages of the simulation. Once you're
> > looking at convergence of the PMF, it's a bit late to decide on it. The
> > choice is a trade-off between the speed of the initial exploration, and
> > non-equilibrium work in that period. In doubt, leave it at its default
> > value.
> >> Is there anything else to check for this particular colvar?
> > Be aware that you won't see a value or zero, or very small values, and
> > that the PMF will rise steeply for values close to zero. That's just a
> > mathematical property of the RMSD function. To make sure to avoid
> > numerical problems due to this singularity, you could set the
> > lowerBoundary to a small but non-zero number.
> > Cheers,
> > Jerome
> Dr. George Patargias
> Postdoctoral Researcher
> Biomedical Research Foundation
> Academy of Athens
> 4, Soranou Ephessiou
> 115 27
> Office: +302106597568
-- Aron Broom M.Sc PhD Student Department of Chemistry University of Waterloo
This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:16 CST