axial stereochemistry search

User 870ab5b546

21-09-2011 17:02:09

I was exploring your new axial stereochemistry search option.  Consider the case shown in the figure.  Here are the search parameters:


INFO: JChemCompare.jsp: searchType = 4 (FULL)
JChemCompare.jsp: setStereoSearch = 4 (STEREO_ENANTIOMER)
JChemCompare.jsp: ignoreDoubleBondStereoMatching = false
JChemCompare.jsp: stereoMatchingModel = 0
JChemCompare.jsp: chargeType = 1
JChemCompare.jsp: radicalType = 1
JChemCompare.jsp: isotopeType = 1
JChemCompare.jsp: valenceType = true
JChemCompare.jsp: considerOddCumuleneStereoMatching = true
JChemCompare.jsp: considerAxialStereoMatching = true
JChemCompare.jsp: considerSynAntiStereoMatching = true
JChemCompare.jsp: bondVagueness = 0
JChemCompare.jsp: orderSensitive = false
JChemCompare.jsp: listAllMatches = false
JChemCompare.jsp: target = [H]C(Cl)=[C@@]=C([#6])C1=CC=CC=C1 |r,c:8,10,t:6|
JChemCompare.jsp: query = [H]C(Cl)=[C@]=C([#6])C1=CC=CC=C1 |r,c:8,10,t:6|
JChemCompare.jsp: searchResult = false

With the STEREO_ENANTIOMER search type, the target should match the query, but it doesn't.


Also, note that if you paste the SMILES strings into Marvin, it ignores the allenic stereochemical information.  It should not.

ChemAxon 42004978e8

26-09-2011 13:31:26

Hi Bob,


From the stereo types: cumulenes, axial stereo configuration and syn-anti stereo only the last is supported in cxsmiles.


Enantiomer search works as if all the centers had an AND label. Enhanced stereo labels only work for tetrahedral stereochemistry.


Bye,


Robert

User 870ab5b546

26-09-2011 14:21:44

It doesn't matter to me why your algorithm is buggy; it only matters that it is.  If you advertise that the STEREO_ENANTIOMER option allows enantiomers to match with one another, then you need to make sure they match.  I suggest that if the STEREO_ENANTIOMER option is turned on, that you look for a [C,N]=C=[C,N] substructure in the target; if you find it, you need to compare both the target and its mirror image to the query.



And there's no good reason why, when pasting a SMILES string that contains stereochemical information (such as [H]C(Cl)=[C@@]=C([#6])C1=CC=CC=C1 |r,c:8,10,t:6|) into Marvin, that information should be lost.  Please fix.


ChemAxon 42004978e8

26-09-2011 18:20:24

Hi Bob,


Our current smiles representation encodes tetrahedral and C/T stereo information. Even though these configurations can be expressed with wedges they are not tetrahedral. 


Enantiomer search is meant for tetrahedral stereo only. 


Bye,


Robert

User 870ab5b546

26-09-2011 18:43:50

Yes, I understand what you have.  I am requesting that you improve it so that it makes more sense from the user's point of view.  A user who sees STEREO_ENANTIOMER is going to expect it to apply to allenes as well as tetrahedral stereochemistry.  And from a chemical point of view, it should.  So please add that capability.

User 870ab5b546

28-09-2011 17:46:18

It would also be good for the StereoisomerPlugin to understand and apply axial stereochemistry.

ChemAxon 42004978e8

07-10-2011 06:42:17

Hi Bob,


Yes, we know about this feature request (https://www.chemaxon.com/forum/ftopic8109.html) we will implement it when it fits to our schedule. 


Regards,


Robert