AMBER Archive (2003)

Subject: RE: doubts

From: Yong Duan (yduan_at_udel.edu)
Date: Fri Mar 21 2003 - 08:30:59 CST


Dear Thenmalar:

My guess is that couple of atoms got too close to each other during the
conjugate gradient minimization phase. Check atom 956, C4' and the atoms
around it.

Conjugate gradient is more aggressive and somewhat more speculative than
steepest descent. As a consequence, it is generally less reliable
(stable) than SD.

By the way, is there any special need for such an elaborate
minimization? 237,700 steps is equivalent to about 500ps. I typically
start with 100-500 steps of minimization and let MD to take care of the
equilibration.

yong

-----Original Message-----
From: thenmalar rathinavelan [mailto:thenmalrr_at_yahoo.co.in]
Sent: Thursday, March 20, 2003 11:27 PM
To: amber_at_heimdal.compchem.ucsf.edu
Subject: doubts

Dear Ambers,
   I have following doubts. That is,
(1) I wish to calcualte the free energy of formation
of a DNA system. For that, I am carrying out steepest
descent (1000cycles) and conjugate gradient
minimization for the convergence of 0.0001 in vaccum.
However, during the minimization following error
occurs.

  NSTEP ENERGY RMS GMAX
   NAME NUMBER
 237700 -2.9549E+02 6.7657E-04 1.1241E-02
    C6 966

 BOND = 23.2027 ANGLE = 197.5915
DIHED = 673.3827
 VDWAALS = -1114.8348 EEL = 3.5336
HBOND = 0.0000
 1-4 VDW = 320.5203 1-4 EEL = -398.8836
CONSTRAINT = 0.0000
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0

   NSTEP ENERGY RMS GMAX
    NAME NUMBER
 237750 1.6690E+07 1.0082E+05 4.0748E+06
    C4' 956

 BOND = 16167904.3232 ANGLE = 19742.2772
DIHED = 1222.2309
 VDWAALS = 492222.7001 EEL = 3.6274
HBOND = 0.0000
 1-4 VDW = 9110.5214 1-4 EEL = -383.8571
CONSTRAINT = 0.0000
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0

   NSTEP ENERGY RMS GMAX
    NAME NUMBER
 237800 2.5064E+13 9.7350E+06 3.2749E+08
    C5 448

 BOND = ************* ANGLE = 34439.3569
DIHED = 1702.4299
 VDWAALS = 17213525.0913 EEL = -68.8777
HBOND = 0.0000
 1-4 VDW = 14025.0206 1-4 EEL = -335.9435
CONSTRAINT = 0.0000
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 1023791 HBPAIR = 0
 NB-update: NPAIRS = 977483 HBPAIR = 0
 
kindly, tell me what could be the reason for this
error. Input I have used is,

 &cntrl
 timlim=999999., idiel=0, dielc=4.0,
 cut=99.0, ntnb=1, nsnb=10, scee=1.2,
 imin=1, maxcyc=9000000, ncyc=1000,
drms=0.0001
 &end

  
(2) I have another basic question regarding prep and
parm.dat files.
    For the dynamics or minimization bond distance,
angle and torsional parameters given in parm
(parm94.dat or parm.96.dat or ....) file are used.
However, we are specifying bond length, angle and
torsional angle values in the prep file also. I like
to know whether prep values be overridden (except
parital charges) by the values specified in the
parm.dat file.
   Please clarify this point also.

Thank you very much in advance.

yours sincerely,
thenmalar

________________________________________________________________________
Missed your favourite TV serial last night? Try the new, Yahoo! TV.
       visit http://in.tv.yahoo.com