Re: [AMBER] Sudden change of energies in min.out files

From: Jason Swails <>
Date: Mon, 22 Sep 2014 10:10:14 -0400

On Mon, 2014-09-22 at 18:15 +0800, maryam azimzadehirani wrote:
> Dear Amber users,
> I have a transition pathway(100 pdbs) from the inactive to active state of
> my protein. After minimizing every structure of the pathway with the below
> file,
> &cntrl
> imin=1,
> igb=5,
> ntx=1,
> ntpr=10,
> ntb=0,
> cut=999,
> rgbmax=999,
> maxcyc=10000, ntmin=1, ncyc=2000

This is not a very "strict" minimization protocol. If you are really
looking for a minimized potential energy surface, I would suggest
setting a fairly strict minimization tolerance (drms -- the default of
1e-4 is probably good) and running the minimization until the
convergence criteria is met (i.e., set maxcyc _very_ large). And the
XMIN minimizer is a much better minimizer than the default one (see the
ntmin=3 option in the manual). It could also be that, for whatever
reason, one of your structures minimized "better" than the others.

> I plotted the energy and ele and vdw terms from the min.out files, the
> plots are attached. in conformation 73 and 74 I have a peak in ele and vdw
> and in energy there is a decrease.How can I know what caused this changes?

Visualizing the structure is a good start. Also, individual components
of the potential is not what I would be comparing here. Minimization
works on the _entire_ potential, not just the electrostatics or van der

> and since the values are for the entire protein, do you have any
> suggestions on how to find out what is happening residue wise?

You can use residue decomposition. Look at the "idecomp" variable on
page 288 of the Amber 14 manual.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Mon Sep 22 2014 - 07:30:05 PDT
Custom Search