[gmx-users] 4-letter residues

[João Henriques]

Dear all,

I have a protein with a phosphorylated threonine and the corresponding
residue name in my FF of interest is THP1. 4-letter residues and PDB files
are a recipe for disaster, but somehow pdb2gmx detects the residue's name
and number perfectly. What happens next is somewhat odd (at least to me).
It now ends in a fatal error regarding OXT atoms. This should not occur
and, in fact, the problem disappears as soon as I use the same exact PDB
but with a regular non-phosphorylated threonine residue instead. The OXT
atoms are still there. Anyone has any idea of what's going on? Should I
edit the FF to use a three letter name instead of THP1?

I'm using gmx 5.1.4.

This is the THP1 residue in my PDB:

ATOM   1171  N   THP1A 160      14.798   3.634  55.810  1.00 28.16
  N
ATOM   1172  CA  THP1A 160      13.457   3.397  56.297  1.00 32.05
  C
ATOM   1173  CB  THP1A 160      13.440   2.416  57.453  1.00 31.31
  C
ATOM   1174  CG2 THP1A 160      12.002   2.135  57.857  1.00 28.13
  C
ATOM   1175  OG1 THP1A 160      14.038   1.189  57.027  1.00 35.99
  O
ATOM   1176  P   THP1A 160      15.501   0.748  57.510  1.00 37.49
  P
ATOM   1177  O1P THP1A 160      15.810  -0.350  56.582  1.00 38.61
  O
ATOM   1178  O2P THP1A 160      16.325   1.972  57.300  1.00 35.09
  O
ATOM   1179  OT  THP1A 160      15.173   0.306  58.890  1.00 40.70
  O
ATOM   1180  C   THP1A 160      12.747   4.659  56.712  1.00 39.18
  C
ATOM   1181  O   THP1A 160      13.202   5.412  57.567  1.00 38.80
  O

This is the error:

Fatal error:
Atom OXT in residue LEU 298 was not found in rtp entry LEU with 19 atoms
while sorting atoms.

Thank you in advance,
Best regards,
João