Wrong angles with pdb2gmx

User discussions
1

GROMACS version: 2020.1
GROMACS modification: No
Hi all,

I am trying to generate a topology file with pdb2gmx for a simple system of a TIP3P water sphere, using CHARMM36 FF (charmm36-jul2020.ff). The command ends without error (gmx pdb2gmx -f sphere_20.pdb -ff charmm36-jul2020 -water tip3p -o sphere_20.gro). However, it writes wrong angles like OH2 H1 H2 and OH2 H2 H1, where, in principle, there should be only one angle H1(2) OH2 H2(1) per water molecule.

Bellow a section of the input pdb file, and the topology generated:

sphere_20.pdb:

ATOM      1  OH2 TIP3    1      32.580  31.090  27.590  0.00  0.00           O
ATOM      2  H1  TIP3    1      32.670  32.030  27.460  0.00  0.00           H
ATOM      3  H2  TIP3    1      31.940  30.810  26.940  0.00  0.00           H
ATOM      4  OH2 TIP3    2      31.510  30.340  31.010  0.00  0.00           O
ATOM      5  H1  TIP3    2      31.470  29.690  31.720  0.00  0.00           H
ATOM      6  H2  TIP3    2      31.190  29.880  30.240  0.00  0.00           H
(...)

topol.top:

[ atoms ]
;   nr       type  resnr residue  atom   cgnr     charge       mass  typeB    chargeB      massB
; residue   1 TIP3 rtp TIP3 q  0.0
     1         OT      1   TIP3    OH2      1     -0.834    15.9994
     2         HT      1   TIP3     H1      2      0.417      1.008
     3         HT      1   TIP3     H2      3      0.417      1.008   ; qtot 0
; residue   2 TIP3 rtp TIP3 q  0.0
     4         OT      2   TIP3    OH2      4     -0.834    15.9994
     5         HT      2   TIP3     H1      5      0.417      1.008
     6         HT      2   TIP3     H2      6      0.417      1.008   ; qtot 0
(...)

[ angles ]
;  ai    aj    ak funct            c0            c1            c2            c3
    2     1     3     5 
    1     2     3     5 
    1     3     2     5 
    5     4     6     5 
    4     5     6     5 
    4     6     5     5 
(...)

Am I missing something. Any help will be appreciated.

Best
-Yasser

2

There are three bonds defined in the .rtp entry for TIP3, including an H1-H2 bond. This is CHARMM convention for a rigid molecule. pdb2gmx generates angles based on bonds, so it will get this wrong, but usually one doesn’t use pdb2gmx to generate a water topology since those are all predefined and you don’t want explicit bonds and angles. TIP3P is rigid and should therefore be defined using SETTLE.

3

Thank you for the reply.

Best.