Successful packing - Chimera/ChimeraX bad PDB record

m3g / packmol

Packmol - Initial configurations for molecular dynamics simulations

http://m3g.github.io/packmol

MIT License

214 stars 48 forks source link

Successful packing - Chimera/ChimeraX bad PDB record #64

Closed ONISniperOne closed 2 months ago

ONISniperOne commented 2 months ago

Seems like everything worked, but opening the output pdb in Chimera yields bad record

Seems to be no problem in VMD.

I think Chimera and VMD has different serialization scheme at the higher numbers. Any solutions?

lmiq commented 2 months ago

If the file opens correctly in VMD, you can try saving it from VMD in another format.

lmiq commented 2 months ago

VMD ignores the atom number from the PDB, thus why it does not complain.

ONISniperOne commented 2 months ago

Maybe I can try to resave as pdb? Should that fix the #ing issue to open in Chimera?

lmiq commented 2 months ago

I don´t know, that depends on what chimera supports. Maybe it is better to save it as "pdbx" (.cif), which supports any number of atoms. But it is to chimera and further software you use to support that or no.

ONISniperOne commented 2 months ago

Update (workaround):

Opened in VMD and saved as .cif as suggested above. Opened the .cif in Chimera/ChimeraX and resaved as pdb file. The bad CONECT record and duplicate serial #ing is fixed. However, some residues can no longer be identified due to the conversion such as CYX (disulfide bonded version of CYS), HIE, and HID. I opened the protein/receptor pdb file used in packmol and the converted system, matchmake the receptor pdb onto the converted, and replaced the protein in the converted system with the one from receptor only pdb from packmol.

lmiq commented 2 months ago

The PDB format is not ideal for dealing with very large structures, and it is only useful in MD simulations because the packages ignore the indexing. Probably that should be done by Chimera as well.

I'm closing the issue, if you have any other question, let me know.