AMBER Archive (2004)

Subject: AMBER: problems in using glycam

From: Wen-Chi Tseng (tsengwc_at_ch.ntust.edu.tw)
Date: Wed Jun 30 2004 - 06:12:22 CDT


I am a newbie to Amber8. When I tried to use the glycam force field, I
encountered a couple of problems.

1. When I loaded the glycam04EP force field into tleap by typing
"source /usr/local/amber8/dat/leap/cmd/leaprc.glycam04EP", I got the
following message.

----- Source: /usr/local/amber8/dat/leap/cmd/leaprc.glycam04EP
----- Source of /usr/local/amber8/dat/leap/cmd/leaprc.glycam04EP done
Log file: ./leap.log
Loading parameters: /usr/local/amber8/dat/leap/parm/glycam04EP.dat
Reading force field mod type file (frcmod)
Unknown keyword: HW OW in parameter file.
Unknown keyword: N NA N2 in parameter file.
Unknown keyword: C C* CA in parameter file.
Unknown keyword: END
 in parameter file.
Unknown keyword: END
 in parameter file.
No mass defined for non-bond atom type: H3 - skipping
Loading Prep file: /usr/local/amber8/dat/leap/prep/glycam04EP.in.

I quited tleap, and reloaded glycam04 force field instead of the
long-pair one. Everything turned out to be normal; the error messages
were gone.
Is there something wrong with the glycam04EP.dat file? If so, please
advise how to fix it.

2. I followed the example in the manual of Amber 8 (p. 26 - 28) to
construct the man9 molecule of the example. But I got some error
messages when I typed "set part1 tail part1.2.16" and "part2 =
sequence { part1 2MA 2MA 1MA }" to set the tail atom of part1 to O3
of VMB and then to join part1 to the long chain.

ERROR: Comparing atoms C3, H2, O2, C1 to atoms C3, H2, C1, O2
!FATAL ERROR----------------------------------------
!FATAL: In file [chirality.c], line 121
!FATAL: Message: Atom C1 is not in the first list
!
!ABORTING.

After examining the sugar structure, I found that the error messages
can be fixed by changing the command "set part1 tail part1.2.16" to
"set part1 tail part1.2.20". But I am not sure if the change is
correct.

I continued to construct the man9 molecule. However, when I checked
the torsions of man9 by following the commands in the example, I got
stuck again.

measureGeom man9.4.1 man9.3.28 man9.3.25 man9.3.20
measureGeom: Argument #2 is type String must be of type: [atom]
usage: measureGeom <atom> <atom> [atom atom]

I checked the man9 molecule using "desc man9.3" and got the following
info. There's no man9.3.25 and man9.3.28 atoms for torsion check.

RESIDUE name: VMA
RESIDUE sequence number: 3
RESIDUE PDB sequence number: 3
Type: undefined
Connection atoms:
 Connect atom 0: A<C1 1>
 Connect atom 1: A<O3 20>
Improper torsions:
Contents:
A<O3 20>
A<O6 19>
A<H62 18>
A<H61 17>
A<C6 16>
A<O5 15>
A<H5 14>
A<C5 13>
A<H4O 12>
A<O4 11>
A<H4 10>
A<C4 9>
A<H3 8>
A<C3 7>
A<H2O 6>
A<O2 5>
A<H2 4>
A<C2 3>
A<H1 2>
A<C1 1>

Does the example in the manual contain typos? If so, can
anyone provide the corrections?

3. I decided to construct a simple sugar, maltose, using glycam04.
Here is what I did and what I got. It seems that I need to define the
torsions and charge of the molecule. Can anyone advise how to do this
and where to get the correct torsion information?

 mal = sequence { 1GA 4GA }
 saveamberparm mal mal.top mal.crd

Checking Unit.
ERROR: The unperturbed charge of the unit: 0.199000 is not integral.
WARNING: The unperturbed charge of the unit: 0.199000 is not zero.

 -- ignoring the error and warning.

Building topology.
Building atom parameters.
Building bond parameters.
Building angle parameters.
Could not find angle parameter: OH - CG - OS
Could not find angle parameter: CG - OH - CG
Building proper torsion parameters.
 ** No torsion terms for HO-OH-CG-H2
 ** No torsion terms for HO-OH-CG-OS
 ** No torsion terms for OH-CG-OS-CG
 ** No torsion terms for H1-CG-OH-CG
 ** No torsion terms for H1-CG-OH-CG
 ** No torsion terms for CG-OH-CG-H2
 ** No torsion terms for CG-OH-CG-CG
 ** No torsion terms for CG-OH-CG-OS
 ** No torsion terms for CG-CG-OH-CG
Building improper torsion parameters.
 total 0 improper torsions applied
Building H-Bond parameters.
Parameter file was not saved.

Thanks

Wenchi

.

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