Re: TIP4P water in AMBER parm7 format fails with 2010-06-07 nightly build

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Sat Jun 12 2010 - 08:41:24 CDT

Hi Hugh,
Aside from being a quick solugion, this is almost certainly the *right*
solution. NAMD doesn't have a hardcoded water geometry, but rather
extracts the rigid water geometry from the equilibrium values present in
the parameter files. This is fine for charmm toppar files, which have to
have all of those values anyway; I hadn't used amber topologies with
NAMD enough to realize that by default you don't get the water angle in
your parm7 file. Let me know if you have further difficulties, but I
would expect things will be fine for you now.
Best,
Peter

On 06/11/2010 04:55 PM, Hugh Heldenbrand wrote:
> Hello Peter and listserv-
>
> I think I may have found a quick solution to this problem, from the
> AMBER file format side rather than the NAMD side.
>
> By default, leap (the AMBER parameter/coordinate generating utility),
> does not include H-O-H angle terms for water molecules in the parm7 file.
>
> You can force these terms to be added to the parm7 file using the
> command:
>
> set default FlexibleWater on
>
> before you add the solvent and save the parameter file. If all of
> these angle terms are written to the parm7 file, namd seems to be able
> to read the file correctly.
>
> At least I am getting a few steps of dynamics when I do it that way
> instead of bombing with an error message. That is as far as I have
> tested.
>
> Thanks,
> Hugh
>
> Peter Freddolino wrote:
>> Hi Hugh,
>> My guess is that there's a difference between the way amber sets up the
>> waters and the way namd expects them, in terms of atom naming or order.
>> If you send me your (compressed) pdb and parm7 files I may be able to
>> find a quick fix.
>> Best,
>> Peter
>>
>> On 06/08/2010 11:30 AM, Hugh Heldenbrand wrote:
>>
>>> Hello-
>>>
>>> I am a new user of namd. I compiled the 2010-06-07 nightly build for
>>> the first time yesterday. I compiled (on an amd64 system) charm-6.1.3
>>> using the intel compiler version 11.1 and ompi version 1.4.1. I
>>> compiled namd using the same intel compiler. I successfully ran the
>>> apoa1 test.
>>>
>>> Now I am trying to run a simulation of a system that I have set up
>>> using AMBER's parmeter/coordinate generating utilities with namd. I
>>> am using the TIP4PEW water model. namd bombs with the following error
>>> message:
>>>
>>> Info: PLACING ATOMS IN PATCHES BY HYDROGEN GROUPS
>>> Info: RIGID BONDS TO HYDROGEN : WATER
>>> Info: ERROR TOLERANCE : 0.0005
>>> Info: MAX ITERATIONS : 100
>>> Info: RIGID WATER USING SETTLE ALGORITHM
>>> Info: RANDOM NUMBER SEED 1276007367
>>> Info: USE HYDROGEN BONDS? NO
>>> Info: Using AMBER format force field!
>>> Info: AMBER PARM FILE ./namd_data/2OUE.parm7
>>> Info: COORDINATE PDB ./namd_data/2OUE_eq81.pdb
>>> Info: Exclusions will be read from PARM file!
>>> Info: SCNB (VDW SCALING) 2
>>> Info: USING ARITHMETIC MEAN TO COMBINE L-J SIGMA PARAMETERS
>>> Reading parm file (./namd_data/2OUE.parm7) ...
>>> PARM file in AMBER 7 format
>>> Warning: Encounter 10-12 H-bond term
>>> Warning: CORRECTION OF ZERO MASS ATOMS TURNED OFF BECAUSE LONE PAIRS
>>> ARE USED
>>> Warning: Found 11254 H-H bonds.
>>> FATAL ERROR: Failed to find water bond lengths
>>>
>>> [0] Stack Traceback:
>>> [0] _Z8namd_diePKc+0x72 [0x4d8922]
>>> [1] _ZN8Molecule17build_atom_statusEv+0x28de [0x90175e]
>>> [2] _ZN8Molecule9read_parmEP4parm+0x1ab7 [0x8fed67]
>>> [3] _ZN8MoleculeC1EP13SimParametersP10ParametersP4parm+0x22a
>>> [0x8fd18a]
>>> [4] _ZN9namdState14configListInitEP10ConfigList+0xf32 [0x918d12]
>>> [5] _ZN9ScriptTcl12Tcl_minimizeEPvP10Tcl_InterpiPPc+0x40b [0x98a04b]
>>> [6] TclInvokeStringCommand+0x91 [0xb45078]
>>> [7] /home/bc2/heldenbr/namd/namd_CVS_Source/Linux-x86_64-icc/namd2
>>> [0xb7aec8]
>>> [8] Tcl_EvalEx+0x176 [0xb7b50b]
>>> [9] Tcl_EvalFile+0x134 [0xb72f14]
>>> [10] _ZN9ScriptTcl3runEPc+0x13 [0x988a03]
>>> [11] main+0x26a [0x4dd2fa]
>>> [12] __libc_start_main+0xea [0x2a964ebaaa]
>>> [13] _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_c+0x3a
>>> [0x4d822a]
>>>
>>> I am not sure what to include from my input file. Some relevant
>>> options that I am using might be:
>>>
>>> waterModel tip4
>>> rigidBonds water
>>> amber on
>>> useGroupPressure yes
>>>
>>> If necessary, I can try to put my input files in their entirety
>>> online. I should note that in addition to the problem reading my
>>> water model, I don't think that the warning about encountering a 10-12
>>> term is correct, the forcefield that I am using does not include those
>>> terms (maybe it encountered one that was 0?)
>>>
>>> Any ideas? Thanks in advance.
>>>
>>> -Hugh Heldenbrand
>>> Dept. of Chemistry,
>>> U of MN
>>>
>>
>>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:55:52 CST