[gmx-users] CGenff parameters in gromacs but missing in charmm package

[Justin Lemkul]

On 4/17/18 11:24 AM, Fabian Keller wrote:
> Dear Justin,
>
>> On 4/17/18 3:34 AM, Fabian Keller wrote:
>>> Dear all,
>>> I am using the CGenFF/paramchem framework for a new molecule (an imidazolium cation derivative with methyl groups on both N atoms).
>>> I faced a known problem going from CHARMM package to Gromacs with parameters already existing in CHARMM for Gromacs. I assume that this is due to paramchem only being implemented for CGenFF3.0.1 and Gromacs using version 4.1.
>> That's precisely your problem. The CGenFF server is using an old
>> version, and therefore it finds bonded parameters that need to be
>> guessed by analogy. In fact, CGenFF 4.0 has these parameters and they
>> are the values you should use. So for all of the parameters that are
>> being overridden, remove them from your .prm file before #including it
>> in the topology. The optimized parameters in CGenFF 4.0 are what you
>> want, and they're all already there in ffbonded.itp.
> Okay thanks, this is what I did. I just was not sure wether the approach was appropriate.
>
>
>>> I wanted now to check if the parameters exist in the CGenFF for CHARMM and they don?t!
>> I'm finding them. It's impossible for grompp to say something is in
>> ffbonded.itp if it isn't. Be sure you're considering forward and reverse
>> orders of atom types when you're searching for them.
> Yes, sure you find them in the charmm36-jul2017.ff/ffbonded.itp, but they are missing in the CGenFF version for CHARMM the toppar_c36_jul17/par_all36_cgenff.prm file (both the latest release). So why is there a discrepancy, or is it even a discrepancy or rather something I am missing here?
>

The GROMACS files come directly from the CHARMM files via a Python 
script. There's no secret magic. If it's in GROMACS, it's in CHARMM. 
You're either looking in the wrong place or you're trying to match the 
wrong string.

-Justin

>>> How is it possible that the parameters exist in the Gromacs version of CHARMM force-field (http://mackerell.umaryland.edu/download.php?filename=CHARMM_ff_params_files/charmm36-jul2017.ff.tgz <http://mackerell.umaryland.edu/download.php?filename=CHARMM_ff_params_files/charmm36-jul2017.ff.tgz>) while they are missing in the CHARMM FF version (http://mackerell.umaryland.edu/download.php?filename=CHARMM_ff_params_files/toppar_c36_jul17.tgz <http://mackerell.umaryland.edu/download.php?filename=CHARMM_ff_params_files/toppar_c36_jul17.tgz>)?
>>> Why are the dihedrals considered duplicate, when the atomtypes clearly are different? Is this due to CHARMM handling dihedral potentials differently to Gromacs?
>>>
>>> Following the link you find all used parameter and force field files (https://www.dropbox.com/sh/m89q64g0pae6wrf/AAAx1aNtJ1TZKNDFgUo2oSu7a?dl=0 <https://www.dropbox.com/sh/m89q64g0pae6wrf/AAAx1aNtJ1TZKNDFgUo2oSu7a?dl=0>) and in the following the error output.
>>> Thank you very much for all hints and help!
>>> Best wishes,
>>> Fabian
>>>
>>>
>>>
>>>
>>>
>>>
>>> WARNING 1 [file chim.prm, line 4]:
>>>    Overriding Bond parameters.
>>>
>>>    old:                                          0.1468 276144 0.1468 276144
>>>    new: CG334   NG2R52     1   0.14530000    251040.00
>>>
>>>
>>> WARNING 2 [file chim.prm, line 11]:
>>>    Overriding U-B parameters.
>>>
>>>    old:                                          107 351.456 0 0 107 351.456 0 0
>>>    new: NG2R52    CG334     HGA3     5   110.100000   351.456000   0.00000000         0.00
>>>
>>>
>>> WARNING 3 [file chim.prm, line 13]:
>>>    Overriding U-B parameters.
>>>
>>>    old:                                          126.3 368.192 0.243 19246.4 126.3 368.192 0.243 19246.4
>>>    new: CG2R53   NG2R52    CG334     5   127.100000   209.200000   0.20900000     12552.00
>>>
>>>
>>> ERROR 1 [file chim.prm, line 25]:
>>>    Encountered a second block of parameters for dihedral type 9 for the same
>>>    atoms, with either different parameters and/or the first block has
>>>    multiple lines. This is not supported.
>>>
>>>
>>> ERROR 2 [file chim.prm, line 26]:
>>>    Encountered a second block of parameters for dihedral type 9 for the same
>>>    atoms, with either different parameters and/or the first block has
>>>    multiple lines. This is not supported.
>
> Thanks a lot,
> Fabian

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

==================================================