certain meso structures

User 870ab5b546

02-05-2005 18:00:47

The following two structures are called identical by JChem 3.0.10, even though they're not:





Code:






  Marvin  05020513552D





  8  7  0  0  0  0            999 V2000


  -10.6875    3.3438    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -7.8296    3.3438    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -9.9730    3.7563    0.0000 C   0  0  1  0  0  0  0  0  0  0  0  0


   -9.9730    4.5813    0.0000 Br  0  0  0  0  0  0  0  0  0  0  0  0


   -9.2586    3.3438    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -9.2586    2.5188    0.0000 Br  0  0  0  0  0  0  0  0  0  0  0  0


   -8.5441    3.7563    0.0000 C   0  0  1  0  0  0  0  0  0  0  0  0


   -8.5441    4.5813    0.0000 Br  0  0  0  0  0  0  0  0  0  0  0  0


  1  3  1  0  0  0  0


  3  5  1  0  0  0  0


  5  7  1  0  0  0  0


  7  2  1  0  0  0  0


  3  4  1  1  0  0  0


  5  6  1  1  0  0  0


  7  8  1  1  0  0  0


M  END











  Marvin  05020513572D





  8  7  0  0  0  0            999 V2000


  -10.6875    3.3438    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -7.8296    3.3438    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -9.9730    3.7563    0.0000 C   0  0  1  0  0  0  0  0  0  0  0  0


   -9.9730    4.5813    0.0000 Br  0  0  0  0  0  0  0  0  0  0  0  0


   -9.2586    3.3438    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -9.2586    2.5188    0.0000 Br  0  0  0  0  0  0  0  0  0  0  0  0


   -8.5441    3.7563    0.0000 C   0  0  1  0  0  0  0  0  0  0  0  0


   -8.5441    4.5813    0.0000 Br  0  0  0  0  0  0  0  0  0  0  0  0


  1  3  1  0  0  0  0


  3  5  1  0  0  0  0


  5  7  1  0  0  0  0


  7  2  1  0  0  0  0


  3  4  1  1  0  0  0


  5  6  1  6  0  0  0


  7  8  1  1  0  0  0


M  END








Note how no parity is shown for C(5), even though it is a stereocenter. Can you help?

ChemAxon 9c0afc9aaf

02-05-2005 18:39:08

Hi,





The mentioned atom cannot be a stereo center, because it doesn't have 4 different ligands, just three (including the implicit hydrogen).





This is because the two connected carbon atoms cannot be distinguished by topology (in other words they have the same graph-invariant value).





Best regards,





Szilard

User 870ab5b546

02-05-2005 18:48:55

Szilard wrote:
Hi,





The mentioned atom cannot be a stereo center, because it doesn't have 4 different ligands, just three (including the implicit hydrogen).


Szilard
But it is a stereocenter. C(2) and C(4) are different from one another: one is R, the other is S. So the central C is attached to four different ligands. In one compound the stereochemistry is denoted as r, and in the other as s. (Note the lower case, because both stereoisomers are achiral.)

ChemAxon 9c0afc9aaf

03-05-2005 06:35:05

Hi,





It seems I was wrong about this issue, sorry.





One of my colleagues will answer this question soon.





Best regards,





Szilard

ChemAxon a3d59b832c

03-05-2005 07:00:01

Well, it still seems that Szilard got close to the root of the problem anyway.





The main problem is that the graphinvariant calculation currently does not include chirality information to avoid a combinatorial explosion. For this reason the middle atom seems to have no parity information as Szilard mentioned. This is what affects both R/S display and searching amongst other things.





We will try to think out a good solution without slowing everything down.





Best regards,





Szabolcs

ChemAxon a3d59b832c

04-05-2005 13:31:13

Bob,





Both the searching and chirality display of these molecules will be fixed in the next major releases. (Marvin 4.0 and JChem 3.1.) You can sign up for the release notification mailing lists by watching these topics:





http://www.chemaxon.hu/forum/ftopic290.html





http://www.chemaxon.hu/forum/ftopic287.html





Best regards,


Szabolcs

User 870ab5b546

04-05-2005 14:12:19

Good news. Thanks. I hope you were able to find a good solution that doesn't slow everything else down.

ChemAxon a3d59b832c

04-05-2005 14:21:42

bobgr wrote:
Good news. Thanks. I hope you were able to find a good solution that doesn't slow everything else down.
Yes, we were, but we are still working on it.





Regards,


Szabolcs

User 870ab5b546

07-10-2005 18:46:39

Hi,





I know this bug was fixed in JChem 3.1. JChem 3.0.15 was released after JChem 3.1 was released. Is the bug fixed in JChem 3.0.15, too?





-- Bob

ChemAxon a3d59b832c

07-10-2005 20:29:34

No, this fix is only available in the 3.1.* releases.





Best regards





Szabolcs