Combining Markush enumeration of stereoisomer expansion

User 99339bfc89

05-10-2011 12:08:37

The Markush enumeration tools in ChemAxon V 5.6, which I have been accessing through Marvin Sketch, are very powerful. I haven't yet, however, been able to figure out how to enumerate all of the different stereoisomers of a molecule which contains other Markush definitions. My first thought would have been to take my Markush molecule and assign one of the bonds coming off of a tetrahedral carbon a zigzag line ( which in Marvin Sketch is labeled as bond type "Single Up or Down" ), but the enumerator then simply carries that zigzag line through to all the enumerated possibilities, instead of expanding it.  To be sure, if I define a simpler structure with no Markush expansions ( ex. no R-groups, or no frequency variation groups, no positional variation bonds, etc.), then I can use the Tools>Isomers> Stereoisomers menu item to request an isomer expansion, and that works fine, but then I've lost the ability to do the Markush enumeration that I came for in the first place. Is there a way to combine these two capabilities, so that I can perform a Markush enumeration AND generate all of the stereoisomers of the resulting molecules at the same time? 

Thanks very much for any help.  -Ben

ChemAxon 42004978e8

06-10-2011 09:34:08



Unfortunatelly this is not possible currently.

Of course trivial and tedious work-arounds exist in which the given stereo center is multiplied in an Rgroup definition with different directions.



ChemAxon a3d59b832c

07-10-2011 14:28:57

Well, not in Marvin Sketch, but it would be possible to pipe two cxcalc commands or execute them one after the other. The first could do Markush enumeration, and the second could do the stereoisomer generation.

Something like this:

cxcalc markush.mrv -o markush_output.sdf enumerations
cxcalc markush_output.sdf -o final_output.sdf stereoisomers

On Unix/Linux systems piping works too:

cxcalc markush.mrv enumerations | cxcalc -o final_output.sdf stereoisomers



A KNIME or Pipeline Pilot protocol could also do the same.


Best regards,


ChemAxon 049d8dcc29

12-10-2011 02:43:29

I have attached an example workflow in KNIME. The input is a Markush structure with two unspecified chiral centers. The workflow first enumerates the Markush structure, then enumerates the stereoisomer. Please test and see if it answers your question. 


-David Deng