[gmx-users] OPLSAA parameters

[Justin A. Lemkul]

Elisabeth wrote:
> Hello to all,
> 
> I am looking for bonded and nonbonded parameters for C H atoms in 
> saturated hydrocarbons.
> 
> atom types are:
> Carbon atoms opls_135      opls_136     
> H  opls_140    
> 
>  opls_135   12.01100  ; alkane CH3
>  opls_136   12.01100  ; alkane CH2
>  opls_140    1.00800  ; alkane H.
> 
> Please help me out with the following:
> 
> - I am wondering if I am selecting the correct values, for C-C and C-H 
> bonds from ffoplsaabon.itp.
> 

The C-C bond is not correct for alkanes.  It is for the C-C bond of oxalic acid, 
as the comment indicates.  In ffnonbonded.itp, you can see how opls_* types are 
mapped to different bonded atom types (for instance, opls_135/136 are CT).  Note 
that in the topology, you do not have to explicitly define values when using 
OPLS-AA; any bond between two defined bonded atom types will be automatically 
detected and used.

> C     C       1    0.15100   292880.0   ; wlj oxalic acid, etc.
> HC    C       1    0.10900   284512.0   ; wlj 7/96
> 
> - what does wlj mean?
> 

Probably someone's initials, which occur frequently to indicate either the 
source of the parameters or the individual that added them.  In this case, my 
guess would be Jorgensen is the referenced individual here.

> - what are the units?
> 

This is in the manual.

> 
> Angle
> 
>  CA     CA     CZ      1   120.000    585.760   ; wlj
>   CA     CA     CR      1   120.000    527.184   ;
>   CA     CA     CX      1   120.000    711.280   ;
>  C      CT     Cl      1   109.800    577.392   ; wlj
> 
> - I can not distinguish CA CZ .... What are they referring to?
> 

See the above tip about converting atom types in ffnonbonded.itp.

> [ dihedraltypes ]
> 
>  C      C      CT     HC      3      0.17782   0.53346   0.00000  
> -0.71128   0.00000   0.00000 ; dicarbonyls BMC 8,1881(2000)
> 
> - Is this what I need for oplsaa 135 136 140?
> 

Probably not.  The comment indicates the dihedral is for dicarbonyls.  The 
appropriate dihedral is:

   CT     CT     CT     HC      3      0.62760   1.88280   0.00000  -2.51040 
0.00000   0.00000 ; hydrocarbon all-atom

Note how the comment indicates it is for use with hydrocarbons.  Again, like 
bonds as described above, you do not need to explicitly list parameters in the 
topology.  They will be inferred from the atom types.

> ffoplsaanb.itp.
> 
>  opls_135   CT    6     12.01100    -0.180       A    3.50000e-01  
> 2.76144e-01
>  opls_136   CT    6     12.01100    -0.120       A    3.50000e-01  
> 2.76144e-01
> 
>  opls_140   HC    1      1.00800     0.060       A    2.50000e-01  
> 1.25520e-01
> 
> - are these parameters for
> 
>  opls_135   12.01100  ; alkane CH3
>  opls_136   12.01100  ; alkane CH2
>  opls_140    1.00800  ; alkane H.
> 
> ?
> 

Yes.

> Also in usr/local/gromacs/share/gromacs/top/ directory  I can not find 
> ffoplsaanb,itp and ffoplsaabon.itp for 4.5.4 whereas in previous 
> versions these files were accessible. I am using itp files from version 
> 4.0.7 and wanted to make sure the above parameters for oplsa_135 
> oplsa_136 and oplsa_140 have been remained unchanged in 4.5.4. Where can 
> I find these files?
> 

The force field organization has changed, but the content of the force fields 
has not.  This is described in the manual.  All the equivalent files are now in 
the oplsaa.ff subdirectory.  For backwards compatibility, the statement #include 
"ffoplsaa.itp" will indeed call the force field, because you will note that 
ffoplsaa.itp now simply calls the correct force field with #include 
"oplsaa.ff/forcefield.itp."

-Justin

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

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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