[gmx-users] A problem? Attn: authors/developers

[Pradip Kumar Biswas]

# To whom it may concern #

While checking the individual pair of coulomb interactions for the case 
of a water-dimer   (OA=0; HA1=1; HA2=2;  OB=3; HB1=4; HB2=5) with OPLS 
force field in gromacs-3.2.1, I found:

For Coulomb interactions evaluated inside inl1000()  it is calculating 
only the terms
(verified from printing the iinr[] and jjnr[] lists & also running in 
the gdb mode):

(v_04 + v_05) + (v_14 + v_15) + (v_24 + v_25)

Considering that due to bonds the program will not calculate the 
coulomb for (v_01, v_02, v_12) & (v_34, v_35, v_45), the terms: v_03, 
v_13, & v_23 which are the coulomb  interactions  with the 2nd oxygen 
(atom_id=3) are missing from the calculation at this stage: list jjnr[] 
contains atoms 4 & 5; atom-3 is somehow excluded!!

Can anybody please tell why it is so or where I am making the mistake? 
I am new in this business; so I apologize in advance if I am drawing 
your attention for any silly mistake.
The residue type is taken as:

[ TIP3 ]
  [ atoms ]
    OH2          opls_111	-0.834  1
     H1	   opls_112 	 0.417  1
     H2	   opls_112 	 0.417  1
  [ bonds ]
    OH2   H1
    OH2   H2


Thanks in advance.
Pradip,
CSU, Ohio.