AMBER Archive (2004)

Subject: AMBER: Re: a new question

From: Viktor Hornak (hornak_at_csb.sunysb.edu)
Date: Mon Apr 12 2004 - 22:02:35 CDT

Xiaowei,

I am forwarding your question to the list where more people can help you
(and better than I ever could :-). In general if you minimize only with
the purpose of running MD afterwards, lower gradients than what you got
are unnecessary. If you are minimizing for normal mode analysis, you may
want to check minimization methods in NMODE. The message RESTARTED DUE
TO LINMIN FAILURE is quite normal - the minimizer cannot get any
further. Please check the FAQs on amber website where these issues are
described in quite a detail (FAQ topics: Energy minimization, LINMIN).

Good luck,
-Viktor

xiaowei li wrote:

> Hi, sorry for bothering you again. I want to minimize the energy of
> DNA with 5 base pairs. The input file is as following:
> &cntrl
> igb=1, ntb=0, ntpr=25,
>
> cut=10.0, ntr=0,
>
> imin=1, maxcyc=500, ncyc=250,
> &end
> After it is finished, I got following results :
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 500 -3.0474E+03 2.9500E-01 1.3922E+00 C7 206
>
> BOND = 12.1891 ANGLE = 49.4306 DIHED = 230.7235
> VDWAALS = -198.4742 EEL = -1716.4911 EGB =
> -1295.7632
> 1-4 VDW = 86.4206 1-4 EEL = -215.4168 RESTRAINT = 0.0000
>
> But, I think it is better if RMS is about 10^-4. So I increased maxcyc
> to 5000 and ncyc to 2500. But I got following error message in the
> output:
> RESTARTED DUE TO LINMIN FAILURE. After this message, the minimization
> restarts and finally I got RMS is still around 10^-2. Do you have any
> experience about it? Thank you in advance.
>
> Best,
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu