[gmx-users] question about PRODRG

[jztan at mail.shcnc.ac.cn]

Dear gmx-users,

I want to MD simulation using Gromacs with my complexed system (protein +
ligand), but when I deal the ligand with PRODRG, I encountered some
problems.

Firstly, I want to deal the protein using GROMOS 96 force field (ffG43a1,
the offical one). But for the ligand, I am a little confused about it. The
PRODRG Beta version provides two force field (GROMOS87 and 96). But when I
select GROMOS96, the result of Gromacs Topology stilll give the GROMOS87
format, but I am not sure if the values are based on GROMOS96. The results
are just like the followings, but this one is different from the results
that generate from "The Dundee PRODRG2 Server"(including charge and atom
tpye), which should be GROMOS87, right? So my question is: can I only
change the format of the results generated from GROMOS 96, so that it will
be the same format as it in the ffG43a1bon.itp? After I change the format,
I can use it as GROMOS96 force field, can I?


The result generated from GROMOS96, using "PRODRG Beta":
;
;       This file was generated by PRODRG version AA050207.0507
;       PRODRG written by Daan van Aalten and Alexander Schuettelkopf
;

[ moleculetype ]
; Name nrexcl
GPI      3

[ atoms ]
;   nr      type  resnr resid  atom  cgnr   charge     mass
     1        OM     1  GPI      O6     1   -0.735  15.9994
     2         C        1  GPI     C14    1    0.349  12.0110
     3        OM     1  GPI      O5     1   -0.735  15.9994
     4       CH2     1  GPI     C13    1    0.016  14.0270
     5       CH2     1  GPI     C12    1    0.017  14.0270
     6       CH1     1  GPI     C10    1    0.088  13.0190
     7         C        1  GPI      C9     2    0.300  12.0110
     8        OM     1  GPI      O3     2   -0.854  15.9994
     9        OM     1  GPI      O4     2   -0.854  15.9994
    10       CH2    1  GPI      C7     2   -0.010  14.0270
....................

[ bonds ]
; ai  aj  fu    c0, c1, ...
   1   2   1    0.125    418400.0    0.125    418400.0 ;    O6  C14
.............................

But in ffG43a1 force field ( ffG43a1bon.itp), the [bonds] format should
like this:
 [ bonds ]
 ;ai  aj  fu    c0          c1
 1     2     2 gb_26 (here is differenct)
..............................



The result generated from GROMOS87, using "The Dundee PRODRG2 Server":
;
;       This file was generated by PRODRG version 050728.0532
;
[ moleculetype ]
; Name nrexcl
GPI      3

[ atoms ]
;   nr      type  resnr resid  atom  cgnr   charge     mass
     1      CR61     1  GPI      C4     1   -0.075  13.0190
     2      CR61     1  GPI      C2     1   -0.075  13.0190
     3        CB       1  GPI      C1     1   -0.055  12.0110
     4        CL       1  GPI       I        1    0.280 126.9045
     5      CR61     1  GPI      C3     1   -0.075  13.0190
     6      CR61     1  GPI      C5     2   -0.083  13.0190
     7        CB       1  GPI      C8     2   -0.062  12.0110
     8      CR61     1  GPI      C6     2   -0.048  13.0190
     9         P         1  GPI      P1      2    1.145  30.9738
    10        OA     1  GPI      O2      2   -0.952  15.9994
..........................

[ bonds ]
; ai  aj  fu    c0, c1, ...
   1   2   1    0.139    418400.0    0.139    418400.0 ;    C4   C2
......................


Secondly,  how can I deal with the charge? PRODRG provides "full" and
"reduced" charges. I am sorry for confusing about this. Is "full" same as
"formal"  charges?  Does this charge generate from QM calculation or some
other force field? Can I use this charges? And also I try to use Gaussian
CHelpG charges, but my ligand contain iodine atom, so I can just use the
basis set 3-21G, I think this basis set is quite small. I only know the
bigger basis set, the lower energy we can get, but I am not sure if it
effect the charges obviously, or hugely?


Finally, my ligand has a I atom, but after calculating using PRODRG, the
result is that the atom type change to "BR" in [atom], but in [bonds], it
is still a "I". So I am not sure if the c0 and c1 values are for "BR" or
for "I"? The results are the followings:

[ atoms ]
;   nr      type  resnr resid  atom  cgnr   charge     mass
    ................
    20         C     1  GPI       C1     4   -0.033  12.0110
    21        BR     1  GPI       I      4    0.169   126.9045
    22       CR1     1  GPI       C3     4   -0.026   12.0110
    23        HC     1  GPI       H3     4   -0.008   1.0080
   .....................

[ bonds ]
; ai  aj  fu    c0, c1, ...
   20  21   2    0.209   2928800.0    0.209   2928800.0 ;    C1    I
 ..........................


Sorry for such a long mail with these questions. Please give me some
advice if you have experience about this. Thanks a lot.


Jinzhi Tan
2002-10-12