Loss of double and triple bonds when saving as .pdb

User 96b281650a

20-09-2012 10:54:58

Im trying to add azide and alkyne groups to a drug compound but when I save it as a .pdb file the triple bonds are all removed.


Can anyone please tell me why this is happening and how I can stop it.


Thank you!

ChemAxon 2cd598e7ad

20-09-2012 14:20:23

Hi,


We are investigating the problem and answer you as soon as we found the cause.


Regards, Domi

ChemAxon 2cd598e7ad

24-09-2012 08:45:23

Hi,


Unfortunately PDB format does not store bond types, just atoms and coordinates. Therefore bond types are estimated from bond lengths. Consequently, if you clean your molecule in 2D, then all bondlengths will be the same, and so reopen makes all bonds single.


There are possible solutions which are explained here, but mainly our suggestion is to use an other format for example mrv or mol formats if storing bond types is important for you.


Solution 1: If you clean the molecule in 3D then the bond lengths will be calculated according to the bond types, that means double and triple bonds will be shorter.
Solution 2: This can be reached also by manually drawing the bonds shorter.


Regards, Domi

User 96b281650a

26-09-2012 21:09:20

Sorted!


Converted the mol format to pdb with OpenBabel and use the --gen3D function. All bonds are there.


Thank you!

User 96b281650a

26-09-2012 21:09:46

Sorted!


Converted the mol format to pdb with OpenBabel and use the --gen3D function. All bonds are there.


Thank you!

ChemAxon 25dcd765a3

26-09-2012 21:33:23

So you originally have a molfile 2D structure and you modify it in the editor and save as pdb file format?


This is the case?

ChemAxon 25dcd765a3

26-09-2012 21:40:37

I realized that the origin of the problem is that molecules when saved to pdb file should be always converted to 3D before saving. So no need to use OpenBabel, just clean the molecule to 3D before saving (press "ctrl + 3"). We will solve this issue. Thank you for the report.

User 96b281650a

26-09-2012 21:41:26

Not necessarily...this was just the method that worked.


Originally tried to edit the pdb file directly but found that converting the pdb to mol and then editing the mol and converting back to pdb with 3D coordinates worked best.