[gmx-users] The energy minimization....

Alan Dodd anoddlad at yahoo.com
Fri Aug 17 16:00:07 CEST 2007

The end structure is the same as the start because gromacs cannot find a lower energy structure than the initial one.  This indicates something severely wrong, either with the topology or starting structure.  You could get a better idea of what's going on by telling gromacs to output structures every step in the minimisation, and examining these so you can actually observe what is happening to your ATP.
I've occasionally seen similar results from minimisation, but only after doing something REALLY bad, like precisely overlaying two molecules on the same coordinates.  Debugging something like this is probably something only you are going to be able to do, as there are just so many potential ways a suitably, uh, imaginative user can mess things up.  

----- Original Message ----
From: MoJie Duan <mjduan at smail.hust.edu.cn>
To: gmx-users at gromacs.org
Sent: Friday, August 17, 2007 11:15:40 AM
Subject: Re: [gmx-users] The energy minimization....

Hi, Mark:
I have done the energy minimization and simulation of ATP in vacuum individual ( Maybe you have suggested me to do this yesterday, but actually I did not understand it, and just do this in solution).
There are following problems:
1. in the energy minimization, the potential energy is positive and in 14th step, it's potential energy is "nan". But the ".gro" outfile of "mdrun" is just the same as the original file (i.e. the .gro file before minimization), the return messages of mdrun is (there are not any warning):

Getting Loaded...
Reading file ATP.2_min.tpr, VERSION 3.3.1 (single precision)
Loaded with Money

Back Off! I just backed up ATP.2_minrun.edr to ./#ATP.2_minrun.edr.1#
Steepest Descents:
   Tolerance (Fmax)   =  1.00000e+00
   Number of steps    =     & #160;    200
Step=    0, Dmax= 1.0e-02 nm, Epot=  1.06678e+05 Fmax= 3.63368e+06, atom= 26
Step=   14, Dmax= 1.2e-06 nm, Epot=          nan Fmax= 3.63313e+06, 
atom= 26
Stepsize too small, or no change in energy.
Converged to machine precision,
but not to the requested precision Fmax < 1

Double precision normally gives you higher accuracy.

writing lowest energy coordinates.

Back Off! I just backed up ATP.2_minrun.gro to ./#ATP.2_minrun.gro.1#

Steepest Descents converged to machine precision in 15 steps,
but did not reach the requested Fmax < 1.
Potential Energy  =  1.0667836e+05
Maximum for ce     =  3.6336775e+06 on atom 26
Norm of force     =            nan

2. in the full MD simulation, the "warning" coming, the messages is:

Step 0, time 0 (ps)  LINCS WARNING
relative constraint deviation after LINCS:
max 0.881911 (between atoms 27 and 28) rms nan
bonds that rotated more than 30 degrees:
 atom 1 atom 2  angle  previous, current, constraint length
     27     28   90.0    0.1610   0.3030      0.1610
Wrote pdb files with previous and current coordinates
step 2490, remaining runtime:     0 s   & #160;           Writing final coordinates.

Back Off! I just backed up ATP.2_runout.gro to ./#ATP.2_runout.gro.1#
step 2500, remaining runtime:     0 s               

And in the ".gro" file after this step, the coordinates of all atoms are "nan". So it means there are crash in the structure? Is the crash between the atom 27 and 28? How to modify the structure file make it normal?
Thank you very much!


>MoJie Duan wrote:
> > >So look at your structures like I said last time! I'm not her e to give
> > >my valuable time giving free advice in order to have it ignored...
> > Thank you very much for your kindness and patience. Maybe sometimes my 
> > questions seems to be silly and boring, my knowledge about GROMACS is 
> >really lack, sorry.....
> >Actually, I really cannot understand what you said yesterday. Did you 
> >mean is there any difference between the atom coordinates of ATP before- 
> >and after- minimization?

> There would normally be some differences visible. If your topology was
> badly broken, then you would usually see where it was broken.

> >I found there are not any difference between 
> >these two structure.
> >(There are also not any obvious collision between atoms of ATP when 
> >represent it by Rasmol)

> OK, so that means your structure is in a flat area of the potential
> surface defined by your topology. If the topology is sound, then you're
> in business.

> My first recommendation was to minimize and/or equilibrate these
> structures on their own, and now I suggest doing them also in solvent.
> This will help you eliminate sources of problems and guide you to 
> what the real problem is. Divide and conquer...

> That's fine, then.

> OK, so here's your problem. Work out what's breaking and why. Read the
> error messages and look at the structures. Understand what each of  > your .mdp file options does.


Building a website is a piece of cake. Yahoo! Small Business gives you all the tools to get online.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20070817/3a2ac885/attachment.html>

More information about the gromacs.org_gmx-users mailing list