AMBER Archive (2008)

Subject: Re: AMBER: Problem with: Increase lastrst in the &cntrl namelist

From: rdauria_at_uci.edu
Date: Wed Jan 16 2008 - 12:19:27 CST

Hello,

Thank you very much for your reply. Sorry for not having been too specific
(that was at the end of a frustrating evening). Here are the set of
commands that I issued (after having loaded all of the needed libraries
and frocmoc files) in tleap to generate the coordinate and the parameter
(and pdb also) files (by the way I am using Amber 8):

sys = loadpdb ions.pdb
solvateBox sys POL3BOX 3.2 iso .7
#set sys box {16.000 16.000 16.000}
saveamberparmpol sys test.parm7 test_xyz.rec
savepdb sys fromtleap.pdb

where ions.pdb is:

CRYST1 16.000 16.000 14.000 90.00 90.00 90.00 P 1 1
ATOM 1 Na+ Na+ X 1 3.967 11.883 9.163 1.00 0.00
ATOM 2 Na+ Na+ X 2 6.321 10.754 6.275 1.00 0.00
ATOM 3 Cl- Cl- X 3 5.197 9.567 8.521 1.00 0.00
ATOM 4 Cl- Cl- X 4 6.000 13.178 7.865 1.00 0.00
END

Note that even when I comment out the set sys box and leave the one
generated from tleap I still get the same error, namely:

...
Ewald parameters:
     verbose = 0, ew_type = 0, nbflag = 1, use_pme =
   1
     vdwmeth = 1, eedmeth = 1, netfrc = 1
     Box X = 16.911 Box Y = 17.503 Box Z = 17.586
     Alpha = 90.000 Beta = 90.000 Gamma = 90.000
     NFFT1 = 16 NFFT2 = 18 NFFT3 = 18
     Cutoff= 6.000 Tol =0.100E-04
     Ewald Coefficient = 0.47291
     Interpolation order = 4

--------------------------------------------------------------------------------
   3. ATOMIC COORDINATES AND VELOCITIES
--------------------------------------------------------------------------------

 begin time read from input coords = 0.000 ps

 Number of triangulated 3-point waters found: 88
| Atom division among processors:
| 0 268
| Running AMBER/MPI version on 1 nodes

     Sum of charges from parm topology file = 0.00000000
     Forcing neutrality...

--------------------------------------------------------------------------------
   4. RESULTS
--------------------------------------------------------------------------------

 Exceeding lastrst in get_stack
   lastrst = 21222
   top_stk= 0
   isize = 21281
   request= 21281
  Increase lastrst in the &cntrl namelist

Please note that the code was totally insensitive to augmenting lastrst in
the input file, which was:

# min
 &cntrl
        imin = 1,

        maxcyc = 5000, dx0=0.5,

        ntpr = 10, ntwr = 500,
        ntf = 1, ntc = 1, ntb = 1,
        cut = 6.0,
        lastist = 1500,
        lastrst = 24000,
 &end

The problem is probably only with the cutoff of 6. and if fact I just
tried to put the cutoff at 8 and skinnb at 0 (bear with me I only need to
set up a small system with the right density and no bad contacts to then
run it in a very different code) and guess what? At least the minimization
worked!!! But now it looks like I am back at exploding bubbles...

A suggestion: why not writing on the manual that 8 is not just the default
value of the variable CUT but even its possible minimum value (I
understand that in general people don't use amber to do very small systems
but just in case they needed to set up one in it)?

Thanks a lot :)

Raffaella.

On Wed, January 16, 2008 8:46 am, David A. Case wrote:
> On Tue, Jan 15, 2008, rdauria_at_uci.edu wrote:
>>
>> I need to set up a small system with ~80 water molecules and few ions,
>> therefore I have solvated 4 simple ions (2 sodium and 2 chloride) in a
>> POL3 water box with the command solvatedBox. I created with tleap the
>> parm7, the coordinate file and also the pdb file with savepdb command. I
>> used setBox to set the PDB. My final box is a 16 Angs cube. I have set
>> the
>> cutoff to 6. Angs.
>
> I can't help with the specific problem, but Amber can't handle such small
> systems very well. Basically, the box size must be at least twice (cut +
> skinnb). Given than an 8 Ang. cutoff is a practical minimum with the
> current
> codes, you are stuck with using larger systems.
>
> You should also indicate exactly how you created the system, since that is
> probably related to the actual error. There should be no need to use a
> setBox
> command if you also used solvateBox, and I suspect that the origin of your
> problem lies in that direction. But we need more details to provide help.
>
> [Generally, to everyone: don't just say things like "I used setBox to set
> the
> pdb". Please give the *exact* sequence of commands. Generally, when
> things
> go wrong, it's because of details.]
>
> ...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

________________________________
Raffaella D'Auria, PhD
Department of Chemistry
3218 Natural Sciences 1
University of California, Irvine
Irvine, CA, 92697-2025
ph. n.: 949-824-9921
e_mail: rdauria_at_uci.edu
________________________________

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