[gmx-developers] Re: [gmx-users] ffG43a1.rtp file error

[Ken Rotondi]

Rob,

Thanks for the clarification. However, the problem is that pdb2gmx does 
not assign a preprocessor code in the .top file. The .rtp file does 
have

    -O     P   O5*   C5*     gd_11
    -O     P   O5*   C5*     gd_9

for the DNA/RNA backbone, and gd_11 & gd_9 are defined in the .itp file.

I realize that I've made an error in description. I said that GROMACS 
could not assign the  O1P-P-O5*-C5* proper dihedral. Actually it cannot 
assign the proper dihedral for O (from the 3' end of the previous 
nucleotide) -P-O5*-C5*. I assume that is what is being described by the 
"-O" in the .rtp description of the dihedral.

I initially worried that the problem was that GROMACS was getting 
confused by having two names for one atom, -O when referring to the 
dihedral with the prior nucleotide, and O3* when naming the atom in the 
same nucleotide. This would not seem to be the case, as only one proper 
dihedral is not named and the first two backbone dihedrals are 
described as

;  ai    aj    ak    al   gromos type
    -C    -O     P   O5*     gd_11
    -C    -O     P   O5*     gd_9
    -O     P   O5*   C5*     gd_11
    -O     P   O5*   C5*     gd_9

in the .rtp file.

This is so tantalizingly close, there must be something I'm missing

help, please,

Ken


On Mar 21, 2006, at 12:15 PM, rob yang wrote:

> Hi Ken,
>
>> #define gd_9      0.000       3.14          2
>
> 0.000 = phase shift. it takes on values of 0 or 180.
> 3.14 = force constant in KJ/mol
> 2 = multiplicity of the torsion term
>
> the torsion equation is: Sum of {(Force constant)/2*(1+cos(phase 
> shift)*cos(multiplicity * your angle))}
> it can theoretically go up to 6 terms (multiplicities)
>
> it seems like there are two terms in this particular torsion, so if 
> you are sure that this is indeed your dihedral angle, you can add two 
> lines to your .itp file:
> 2     1     4     5     1 gd_9
> 2     1     4     5     1 gd_11
>
> hope that helps.
>
> Rob
>

Hi All,

I've processed an RNA chain in GROMACS 3.2.1 using the G43a1 ff to .gro 
and .top files.  Upon trying to generate a .tpr file for genion, grompp 
returned an series of warnings leading to a fatal error, eg.:

WARNING 2 [file "rnatest.top", line 605]:
  No default Proper Dih. types, using zeroes

looking at the .top file line 605 I find:

[ dihedrals ]
;  ai    aj    ak    al funct            c0            c1            
c2            c3            c4            c5
    2     1     4     5     1 
    1     4     5     6     1    gd_14

Which indicates that GROMACS does not know how to assign the 
O1P-P-O5*-C5* proper dihedral in the RNA backbone, as this dihedral is 
blank for every nucleotide.

When I look in the G43a1.rtp file I find that there are two listings 
for this dihedral:

    -O     P   O5*   C5*     gd_11
    -O     P   O5*   C5*     gd_9

With the gd_11 and gd_9 pre-processor codes defined in the .itp file as:

#define gd_9      0.000       3.14          2
; -OA-P-        0.75
;
#define gd_11     0.000       1.05          3
; -OA-P-        0.25

It seems to have everything it needs, but does not complete the job. 
What is needed?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 3377 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-developers/attachments/20060321/5176767a/attachment.bin>