Dear David.
Thank you, I will have a try ntmin=2 settings.
As about number of minimization steps I used 10 times less. Increasing the
number to such extend was my naive attempt to resolve the problem with
overlapping atoms.
Thanks for pointing out the "weak" mismatch, that is "my slang", it is
weaker then 500 which I tried before.
Best regards.
Petro
On Thu, Jan 9, 2020 at 3:23 PM David Case <david.case.rutgers.edu> wrote:
> On Thu, Jan 09, 2020, Petro wrote:
>
> >After adding water by addtobox to a pDB and ruining minimization I still
> >have some overlapping water hydrogen. For now I came up with a crude
> >solution and remove these 2-4 waters. But is there any way I can force
> >minimization to solve it?
>
> If you have bad initial water contacts, set ntmin=2 and run 500-1000
> steps of minimization. That will just use Steepest descent, and will
> avoid linmin problems.
>
> However, I suspect that by asking for 100000 (!!) minimization steps,
> you are just wasting computer time: does your energy not converge after
> a few hundred steps?
>
> >here is my minimization parameters :
> >"""
> >simple generalized Born minimization script
> (note that this is not a GB calculation)
> > &cntrl
> > imin=1, ! initiate minimization
> > ntb=1, ! use NVT ensemble
> > maxcyc=100000,
> (consider a value two to three orders of magnitude smaller)
>
> > ntpr =10,
> > cut=12.0, ! cutoff 12 Ang
>
> This is way too big: stick with default of 8.
>
> > ntr=1, ! weak positional restrains
> > iwrap=1
> >/
> >Keep protein fixed with weak restraints
> >10.0
>
> (these are actually fairly strong restriants: you protein atoms will
> not move more that about 0.1 to 0.2 Ang, unless something is really
> wrong.
>
> >The output of minimization repeatedly tells
> > .... RESTARTED DUE TO LINMIN FAILURE ..
>
> This is expected: see above about not requesting so many minimization
> steps.
>
> ...dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jan 09 2020 - 10:30:02 PST
