[gmx-users] Both Steepest and Conjugate didn't converge

[Tsjerk Wassenaar]

Hi Prettina,

We do not receive to many prayers on this user list and nor do I know
whether you want me to praise the lord for being able to help you or simply
in general. But to cut it short, you can just proceed with your md run. You
could have checked the user list archives either with "minimization" and
"converge" or "machine precision". You would have gotten many hits, which
all give the same answer: if you just want to do md, it's enough to have the
worst overlaps and bad contacts relieved a bit.., the md integrator will do
the rest. The negative potential energy indicates that everything is ok.
Only if your intention is to do normal modes analysis you need to have a
completely relaxed structure.

Cheers,

Tsjerk

On 2/8/06, Prettina Anto_Louis <pretti_louis at krify.com> wrote:
>
> Hello gromacs users,
>
>
>            Greetings!
>
>
>                 I am working with a drug-enzyme complex. For my complex, I
> am facing the problem in the energy minimisation step with both steepest
> followed by  conjugate . Both of these integrators, didn't converge with the
> given number of steps. I did the energy minimisation using steepest with the
> following parameters in the em.mdp file:
>
> ==========================em.mdp(steep)=============
>
> cpp                 =  /lib/cpp
> define              =  -DFLEX_SPC
> constraints         =  none
> integrator          =  steep
> nsteps              =  1000
>
> ;
> ;       Energy minimizing stuff
> ;
> emtol               =  2000
> emstep              =  0.01
>
> nstcomm             =  1
> ns_type             =  grid
> rlist               =  1
> rcoulomb            =  1.0
> rvdw                =  1.0
> Tcoupl              =  no
> Pcoupl              =  no
> gen_vel             =  no
>
> ================================================
>
>   I got the follwing result:
>
>
> Stepsize too small (5.71705e-07 nm)Converged to machine precision,
> but not to the requested precision (2000)
>
> Steepest Descents did not converge in 442 steps
>   Potential Energy  = -7.01019e+05
> Maximum force:  9.58302e+03
>
> ============================================
>
>              So, I again performed the energy minimisation using the
> conjugate gradient with the following parameters after the steepest:
>
> ============em.mdp(cg)=========
> cpp                 =  /lib/cpp
> define              =  -DFLEX_SPC
> constraints         =  none
> integrator          =  cg
> nsteps              =  1000
> nstcgsteep          =  1000
> ;
> ;       Energy minimizing stuff
> ;
> emtol               =  1000
> emstep              =  0.01
>
> nstcomm             =  1
> ns_type             =  grid
> rlist               =  1
> rcoulomb            =  1.0
> Pcoupl              =  no
> gen_vel             =  no
> rvdw                =  1.0
> Tcoupl              =  no
>
> =======================
>
>
>       I got the following output for my CG minimisation:
>
>
> Negative number for sqrt encountered (-75353886948327424.000000)
> Terminating minimization
>
> writing lowest energy coordinates.
>
> Back Off! I just backed up em.trr to ./#em.trr.1#
>
> Conjugate Gradients did not converge in 927 steps
>   Potential Energy  = -7.80361e+05
> Maximum force:  1.08202e+04
>
> =============
>
>
>            Please let me know, what is the problem with the energy
> minimisation. Am I going wrong with any of the parameters in the em.mdpfile? Please guide me. Should i proceed with my molecular dynamics with this
> minimised complex coordinates, or should i further minimize the same? Please
> guide me.
>
> Thanks in advance!
>
>
>
>         "praise the lord!"
>
>
> With prayers,
> prettina
>
>
> _____________________________________________________________
> I use Krify Mail - http://mail.krify.com  Get  yourmail at  Krify today!
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>



--

Tsjerk A. Wassenaar, M.Sc.
Groningen Biomolecular Sciences and Biotechnology Institute (GBB)
Dept. of Biophysical Chemistry
University of Groningen
Nijenborgh 4
9747AG Groningen, The Netherlands
+31 50 363 4336
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060208/cc04389b/attachment.html>