How did it fail? Was it a linmin failure? There is a lot on the list about
this. ncyc simply sets the number of "steepest descent" steps that are
taken before conjugate gradient is started. Thus, if you are hitting a
LINMIN failure at some point (basically, the line search can't find the
right direction to minimize, so it backs out and retries, and after 3x it
becomes fatal), increasing ncyc will simply make the minimization converge
too slowly to reach the point where it failed last time (since steepest
descent is slow to converge).
If you're just trying to minimize to relieve bad contacts, you don't need
this many cycles (30000), anyway. Thus, once you reach the linmin failure,
your structure is fine for starting MD anyway. If you're actually looking
for minima, then I would suggest using the xmin minimizer, which is much
better than the built-in minimization routine. Check out ntmin=3 and the
xmin variables in the amber manual.
In the future, though, you should be more detailed about your error
messages. I'm simply guessing at your problem here, since you didn't supply
enough information regarding what went wrong.
Good luck!
Jason
On Fri, Sep 10, 2010 at 3:21 PM, Jiang Yusheng <yusheng0104.gmail.com>wrote:
> Dear all
>
> Here is my input file
>
> minimize system
> &cntrl
> imin=1, irest=0, maxcyc=30000, ncyc=50,
> cut=9.0,
> ntb=0, ntpr = 100, ntr=1
> /
> group input for cart restraints
> 100.0
> RES 32 514
> END
> END
>
> The process failed at 18100 step. Last time I did another work, and I
> adjusted maxcyc and ncyc, the work went through.
>
> Here, I have a question. Is there any relationship between maxcyc and ncyc
> setting? I can set randomly or I can follow some rules.
>
> Thanks
>
> Jason
>
>
>
> --
> Yusheng Jason Jiang
> Department of Chemistry
> Georgia State University
> Atlanta, 30302 GA
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Sep 10 2010 - 13:00:04 PDT
