I'll chime in with my comments here as well.
Have you seen this?
http://ambermd.org/Questions/linmin.html
The issue with a linmin failure is that the minimizer couldn't find the
appropriate direction to move the system in order to reduce the total
energy. The xmin minimizer is much better than the builtin one. To use
this, set ntmin=3. See the manual for more details.
Note that if you're just trying to relax your structure in order to start
dynamics, you structure *should* be good enough by the time you hit a linmin
failure to just start running. I would only advocate using xmin if what you
*really* needed was a minimization as close as possible to a local minimum.
Hope this helps,
Jason
On Mon, Mar 28, 2011 at 5:40 AM, Hugh Heldenbrand <helde010.umn.edu> wrote:
> Hello-
>
> The solution is to restart your minimization not from the beginning, but
> from a restart file that is written for a point in your minimization
> that occurs before the linmin failure.
>
> So, if your failure is occuring at step 653, set maxcyc=500 to get 500
> steps of minimization and a restart file at the end. Then restart your
> minimization using the restart file (not the original coordinate file)
> with maxcyc=4500 to get the rest of the total 5000 steps. If another
> linmin failure occurs, repeat.
>
> -Hugh Heldenbrand
> Graduate Student
> University of Minnesota
>
> On 03/28/2011 07:27 AM, I.Haldoupis wrote:
> > Dear Amber users,
> >
> > As a part of reproducing some published work I need to minimize, using
> > implicit solvation, a ligand in a protein keeping the protein
> > constrained. Unfortunately I get the LINMIN error
> >
> > ".... RESTARTED DUE TO LINMIN FAILURE ...
> >
> > ***** REPEATED LINMIN FAILURE ***** "
> >
> > Looking through the mailing list I see it is a re-occurring issue and
> > that there isn't a clear solution. I tried changing things such as
> > increasing ncyc to 1000, or excluding the gbsa term, or expressing the
> > restrain using the 'restraint_wt' and 'restraintmask' options, however I
> > still get the same error eventually.
> > I tried running it using xmin minimization (ntmin option 3) using the
> > defaults, but terminated early - not sure if that is related to the
> > default convergence criterion for the energy gradient (drms option under
> > xmin).
> >
> > I am just learning Amber and any help would be much appreciated.
> >
> > Please find below my original minimization input file.
> >
> > Minimization with constrained protein
> > &cntrl
> > imin=1, maxcyc=5000, ncyc=100,
> > ntpr=1,
> > cut=16,
> > ntb=0,ntr=1,
> > igb=1,
> > gbsa=1
> > /
> > Hold protein fixed
> > 1000.0
> > RES 1 254
> > END
> > END
> >
> >
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Mar 28 2011 - 09:00:03 PDT
