From: Aron Broom (broomsday_at_gmail.com)
Date: Wed Sep 11 2013 - 00:39:18 CDT
Hi,
I've got a question regarding the recently posted tutorial on calculating
binding free energies that was written up by: James Gumbart, Benoit Roux
and Christophe Chipot. But perhaps others have an answer.
I suspect I'm just being an idiot in some regard, but I'm noticing
something very confusing to me in the *.in files for the geometric route.
It's most evident for the Eular-Theta colvar, named "Theta". So what I see
is that:
1) in the first simulation to be run, where the RMSD is the active bias
being explored and no other biases are on, the Theta colvar has a center at
97.8, but of course no force. My assumption would be the 97.8 was the
value in the starting coordinates. That is from "bound-1.in"
2) in the next simulation to be run, in which the exploration is for
"Theta", the range for the colvar in the "Theta-1.in" file is 65.0 to
95.0. So the 30 degree spread makes sense, but seems odd that it isn't
centered around 97.8, but whatever, maybe I was confused about the 97.8 and
it isn't the starting coordinate value, there is still something more odd
ahead.
3) finally look at the next simulation, where we'll now be restraining the
"Theta" we just explored, and moving our ABF exploration to "Phi", this
being "Phi-1.in". Look at the harmonic restraint for "Theta", it's
centered at 102.0. That seems really weird, since we only told ABF to
explore from 65.0 to 95.0, how is it that we've decided to restrain it to a
value that isn't even in that range?
I would have thought we would restrain it to the PMF minimum we found when
we explored "Theta". I realize that isn't needed, that the whole concept
here is a thermodynamic cycle, so the path doesn't matter, and we can take
a high energy one if we want, but where did 102.0 come from then?
Also, 102.0 is reasonably close to 97.8 from the first file, and it would
seem to make sense that when we explored around the starting value of 97.8
we found the minimum was actually a few degrees off, but that doesn't fit
with the colvars bounds for "Theta" in "Theta-1.in" not including either of
those values.
Am I totally missing how this is supposed to work?
Thanks,
~Aron
-- 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 - 23:21:39 CST