AMBER Archive (2002)

Subject: SHAKE problem

From: Rosalia Pascual (rosalia_at_edurne.iqs.es)
Date: Wed Aug 07 2002 - 10:13:12 CDT

Hi.

I have a problem with SHAKE ( coordinate resetting cannot be ...)
My system is a protein-ligand solvated with a cap of waters of 20 Å
(total charge of the
system is 6).

First, I have minimized it in the following way ( AMBER 7):

Minimization with Cartesian restraints
 &cntrl
    imin=1, maxcyc=5000,
    scee=1.2, igb=0, ntb=0,
    ntr=0, ibelly=0,
    ntmin=1, ncyc=1000, nsnb=25, cut=10,
 &end
 &ewald
    eedmeth=5,
 &end

and then I have run a dynamics for 100 ps:

    ntx=1, ntpr=100, ntwx=100, ntwv=100, ntwe=100,

    ntf=2, ntc=2,
    ntb=0, igb=0, cut=10, scnb=2, scee=1.2,
    ibelly=0, ntr=1,

    nstlim=50000, ntt=1, tautp=0.2,
    dt=0.002,
 &end
 &wt
   type='TEMP0', istep1=0, istep2=10000,
                 value1=0, value2=300.0,
 &end
 &wt
   type='TEMP0', istep1=10000, istep2=50000,
                 value1=300.0, value2=300.0,
 &end
 &wt
   type='END',
 &end
 &ewald
  eedmeth=5,
 &end
 &rst
   iat=0,
 &end
fixed residues (away from the active center)
1000.0
RES 1 2
RES 4 6
RES 65 65
RES 68 86
RES 116 139
RES 172 172
RES 174 187
END
END

After 100 ps, the system is equilibrated ( ETOT constant, RMS
constant...)
Then, I try to run a production ( same as above) for 300 ps.
However I have a SHAKE problem.

I have seen different solutions in similar e-mails, but it doesn`t help
me:

1) VLIMIT is 20

2) The cutoff is smaller ( 10 ) than the diameter of the water cap (20 )

(same
happens if I use cutoff 9 or 8 ).

3) The problem happens 104 ps after the dynamics ( so, maybe is not a
problem
from the initial structure - I have checked it and it seems OK).

4) If I turn off the SHAKE OPTION ( although I have a cap of
WATBOX216 ), temperature raises a lot.

Could anybody help me?
Thanks in advance.