[gmx-users] problems in the topology of nucleotides using ffG43a2.rtp

[Paulo Netz]

The topology of deoxynucleotides in  the ffG43a2.rtp
file is wrong: using this force-field to simulate
DNA causes the deoxynucleotides to remain unconnected,
because no bond between successive residues is defined
in this file. For instance, the bonds in DGUA are defined
as follow:

[ bonds ]
    P   O1P    gb_23
    P   O2P    gb_23
    P   O5*    gb_27
  O5*   C5*    gb_19
  C5*   C4*    gb_25
  C4*   O4*    gb_19
  C4*   C3*    gb_25
  O4*   C1*    gb_19
  C1*    N9    gb_21
  C1*   C2*    gb_25
   N9    C4    gb_9
   N9    C8    gb_9
   C4    N3    gb_11
   C4    C5    gb_15
   N3    C2    gb_11
   C2    N2    gb_8
   C2    N1    gb_16
   N2   H21    gb_2
   N2   H22    gb_2
   N1    H1    gb_2
   N1    C6    gb_16
   C6    O6    gb_4
   C6    C5    gb_15
   C5    N7    gb_9
   N7    C8    gb_9
  C2*   C3*    gb_25
  C3*   O3*    gb_19
  O3*    +N    gb_27

It is easy to see that in the last line the +N
was included by mistake.
This bug can be easily fixed replacing the
+N by +P everywhere in the parameters for
deoxynucleotides within this file.

Paulo A. Netz