Markushcount question

User 677b9c22ff

04-11-2008 09:06:23


I have to files which give me a number of possible structures.

one is coded as R1 = x.y,z the other one as R1 = L{x,y,z}.

In mview I get Markush library size = 390625 (~ 10^6) only for the file with L (see attachement).

Otherwise I get the error:

chemaxon.enumeration.ExpansionException: Circular dependency in order graph.

Cycle atom indexes: 11 9

at chemaxon.enumeration.ExpansionHelper.findExpansionSequence(

at chemaxon.enumeration.ExpansionCounter.createExpansionGraph0(

at chemaxon.enumeration.ExpansionCounter.createExpansionGraph(

at chemaxon.enumeration.ExpansionCounter.countExpansions(


at chemaxon.marvin.plugin.CalculatorPluginManager.runPlugin(

at chemaxon.marvin.plugin.CalculatorPluginManager.runPlugin(

at chemaxon.marvin.plugin.CalculatorPluginManager.runPlugin(

at chemaxon.marvin.plugin.CalculatorPluginManager.callback(

at chemaxon.marvin.plugin.PluginManager.callback(


at Source)

What is wrong? According to the help it seems fine.


Idea: Depth-First

ChemAxon a9ded07333

04-11-2008 09:28:42

Hi Tobias,

Your first Markush structure is not well-formed. The Rgroup that it contains is not an atomlist but an R1 atom itself. It seems to be an atomlist only because it has a label (mrvAlias):




            mrvAlias="H, F, Cl, Br, I"

(You can see the difference by erasing the "mrvAlias=..." line from the file.)

So this is a real circular reference to R1 from inside the R1 definition.

Best regards,


ChemAxon fb166edcbd

04-11-2008 09:46:21

One remark: if you use an atom list as in your -2 files then the R-group notation is not needed, you can just as well replace your R1 atoms by the atom list itself.

Or, you can use an R-group definition by putting the different atoms in different R-group members, as in the attached -3 file.

User 677b9c22ff

04-11-2008 23:47:49

Hello Tamas and Nora,

thank you for your comments and help.