[Open Babel] aromatic - kekule conversion issue

[Nina J. <ni...@ac...>]

Dear All,

My name is Nina Jeliazkova and I’m the lead developer of the Toxtree
application (http://toxtree.sourceforge.net), which relies on Open Babel
for certain format conversions.

We have received user feedback about problems with the following smiles:

c12-c3c(cc(N)cc3)Cc1cccc2

We have traced down the problem to the following info messages,
generated by Open Babel during smi->sdf and smi->smi conversions:

babel ---errorlevel 2 -h -ismi c12-c3c(cc(N)cc3)Cc1cccc2.smi -osmi
c12-c3c(cc(N)cc3)Cc1
cccc2-ob.smi
==============================
*** Open Babel Information  in OpenBabel::OBMol::NewPerceiveKekuleBonds
  Kekulize: Huckel rule not satisfied for molecule

==============================
*** Open Babel Information  in OpenBabel::OBMol::start_kekulize
  Kekulize Error for molecule

1 molecule converted
2 info messages 44 audit log messages

The obtained output is clearly wrong:

C12=C3C(=CC(=CC3)N)CC1C=CC=C2

We have also tried Daylight depict function, available at:

http://www.daylight.com/daycgi/depict/

The later parses and depicts c12-c3c(cc(N)cc3)Cc1cccc2 correctly.

We are well aware that kekule/aromatic standardization efforts are far
from satisfactory so far, however we would like to have your
interpretation of this problem. In particular, is Open Babel behaving as
expected in this particular case?

Kind regards,
Nina