[gmx-users] FF parameter conversion factor from OPLSAA to Gromacs

[Erik Lindahl]

Hi Ruhong,

We know about it, and it should definitely be 4.184. I was a bit stupid  
and picked a value used in Jorgensen's paper instead of the real one.  
It will be fixed for the next release (since it shouldn't make a  
significant difference for things like proteins). Another minor  
correction to OPLS-AA is that you want to change the combination rule  
to "3" in share/top/ffoplsaa.itp.

(We're cross-testing between Amber and OPLS now, and should have both  
in the next version).

Cheers,

Erik


On Friday, Mar 14, 2003, at 13:01 US/Pacific, Ruhong Zhou wrote:

>
> Hi,
>
> I have noticed that the conversion factor used in gromacs from  
> kcal/mol to kJ/mol is
> 4.1868 instead of 4.184 when I was checking the OPLSAA torsion  
> parameters in gromacs
> database. Then I found it is true for the bond stretching parameters  
> as well.
> Is there a particular reason for this 4.1868 J/cal conversion factor,  
> or am I missing something
> here? It is a not a huge difference, but I thought the standard  
> conversion factor should be  
> 1kcal/mol = 4.184 kJ/mol.
>
> Here are two examples. Gromacs converts the OPLSAA torsion parameters  
> to the five-term
> Ryckaert-Bellemans form. The formular is
>
>   C0=V0+V2+0.5(V1+V3)
>  C1=0.5(3*V3-V1)
>  C2=-V2
>  C3=-2V3
>  C4=0
>  C5=0
>
> 1. Torsion CT-CT-CT-CT:
>
> GROMACS:
> CT     CT     CT     CT      3      2.93076  -1.46538   0.20934  
>  -1.67472   0.00000   0.00000 ; hydrocarbon
> OPLSAA (2000):
> CT-CT-CT-CT                          0.000    1.300  -0.050   0.200    
> hydrocarbon
>
> A quick check C2/(-V2) = 0.20934 kJ/mol/(0.05 kcal/mol) = 4.1868    
>  kJ/kcal <======
>
> 2. Bond CT-NT:
>
> GROMACS:
>  CT    NT      1    0.14480   319871.5   ; -idem-
> OPLSAA (2000):
> CT-NT     1.448    382.0
>
> 319871.5 kJ/mol/nm^2 / (2 x 382.kcal/mol/A^2)  = 4.1868    kJ/kcal    
>  <=========
>
> Thanks.
>
> Ruhong
>
> Ruhong Zhou, PhD
> Computational Biology Center  
> IBM Thomas J. Watson Research Center    
> Yorktown Heights,  NY 10598
> Tel: (914) 945-3591, Fax: (914) 945-4104
> http://www.research.ibm.com/people/r/ruhong
>
------------------------------------------------------------------------ 
-----
Erik Lindahl, MSc, PhD     <lindahl at stanford.edu>
D109, Fairchild Building
Dept. Structural Biology, Stanford University School of Medicine
Tel. 650-7250754    Fax. 650-7238464

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 2422 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20030314/ea0d7dac/attachment.bin>