AMBER Archive (2004)

Subject: Re: AMBER: Time to minimize the energy by NMODE

From: Holger Gohlke (gohlke_at_bioinformatik.uni-frankfurt.de)
Date: Wed Apr 14 2004 - 14:40:28 CDT

xiaowei li schrieb:
>
> Hi, I want to calculate the normal frequencies of DNA with 5 base pairs.
> So I take following steps to get it.
> First, using SANDER to minimize the energy, and I got RMS=4.467E-01, and
> sandermin.rst.
> Second, using NMODE(ntrun=4) to minimize sandermin.rst to get RMS is in
> the order of 10^-4.
> As David has emailed me before, it just takes a few hours to finish
> NMODE minimization ( My computer has two 1.53 GHz AMD Athlon K7 MP
> processors with 256KB of cache (per cpu) and 1 GB of RAM).
> However, NMODE minimization has run about 50 hrs on my computer and it
> is not finished yet. Is there anything wrong with my case? Thank you
> for your nice help.
> Following is my input and output file for sander mininization:
> &cntrl
> igb=1, ntb=0, ntpr=50,
>
> cut=999, ntr=0,
>
> imin=1, maxcyc=1000, ncyc=500,
> &end
> output:
> FINAL RESULTS
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 1000 -3.0607E+03 4.4670E-01 2.5204E+00 N3 372
>
> BOND = 12.5946 ANGLE = 50.9148 DIHED =
> 229.0234
> VDWAALS = -200.7355 EEL = -1637.9778 EGB =
> -1382.9776
> 1-4 VDW = 84.9881 1-4 EEL = -216.5779 RESTRAINT =
> 0.0000
>
> Following is the input file for the NMODE minimization:
> &data
> ntrun=4,
> nsave=2,
> ndiag=2,
> cut=10.0,

Try to set the cutoff to 99.0 - atoms that move in and out of the cutoff
region during the minimization may prevent that you reach the specified
gradient.

By the way, I had no problems in reaching a rms gradient of 0.0001
kcal/molA using sander. For a protein with approx. 80 residues, that
only took 1 to 2 hours.

Best regards

Holger

> nprint=1,
> drms=0.0001,
> maxcyc=500,
> scnb=2.0,
> scee=2.0,
> &end
>
> --
> Xiaowei Li
> Department of Electrical and Computer Engineering
> University of Virginia
> 924-3446(O)
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu 

-- 
!!!      Please note new telephone number      !!!
++++++++++++++++++++++++++++++++++++++++++++++++++
Dr. Holger Gohlke

J.W. Goethe-Universität
Fachbereich Biologie und Informatik
Institut für Mikrobiologie	
Marie-Curie-Str. 9              
60439 Frankfurt/Main            
Germany                         

Tel.:  (+49) 69-798-29411; Fax: (+49) 69-798-29826
Email: gohlke_at_bioinformatik.uni-frankfurt.de
URL:   http://www.rz.uni-frankfurt.de/~hgohlke
++++++++++++++++++++++++++++++++++++++++++++++++++
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu