[gmx-developers] gromacs can't recognize residue name with four letters in some cases

[Mark Abraham]

Dongsheng Zhang wrote:
> Dear GMX developers,
> 
> In my pdb file, some residues' names have four letters. When I use
> pdb2gmx, it works fine. The output gro file shows those residues (with
> four letters). When I use editconf -f *gro -o *pdb, gromacs can't
> recognize those residues with four letters. All residues' names in the
> output pdb file have only three letters. For those residues whose name
> have four letters, the last letter has been truncated. Please check it.
> If it is possible, could you please solve it?

The PDB file format is only supposed to permit 3-letter residue names - 
see http://www.rcsb.org/pdb/file_formats/pdb/pdbguide2.2/part_62.html. 
Some PDB readers and writers cheat and allow four-letter residue names, 
but I doubt the best solution is to have editconf join that trend. If 
you need your residue names to be preserved, decide your own mapping 
from four-letter names to unique three-letter names and use commands like

sed -e 's/LYSH/LYH /' -e 's/LYSP/LYP /' old.pdb > new.pdb

Note the space - you need to replace four characters with four 
characters lest some PDB reader depend on the column numbers being right.

Mark