Cannot convert smiles to 3D

Hello all,

I have been trying to convert some molecules from smiles to pdb and xyz and I get the following error.  These molecules are buckyball derivatives and I  have attached a couple of them. I would be really thankful if someone can help me out with it.

 

Multiple anchor bases
Exception in Clean3D. Failed structure:CCCOC(=O)CCC1(c2ccccn2)C23c4c5-c6c2c2c7c8c9c%10c(c%11c4c4c%12c5C5%13c%14c%12c%12c%15c%16c%17c%18c%19c%20c%21c%18c%18c%16c%16c%22c%18c%18c%21c%21c%23c%20c%20c%24c%19c%19c%17c%15c%14c%14c%15c%17c(c6C5%15C%13(CCC(=O)OCCC)c5ccccn5)c2c2c7c(c%23c%20c2c%17c%24c%19%14)c9c%21c%18c%10c%22c%11c4c%12%16)C138
java.lang.IndexOutOfBoundsException
    at chemaxon.marvin.modelling.build.BuildSequence$Build.getAnchorBase(Unknown Source)
    at chemaxon.marvin.modelling.build.BuildSequence$BuildFuse.<init>(Unknown Source)
    at chemaxon.marvin.modelling.build.BuildSequence$BuildFuse.<init>(Unknown Source)
    at chemaxon.marvin.modelling.build.BuildSequence$BuildFuse.<init>(Unknown Source)
    at chemaxon.marvin.modelling.build.BuildSequence$BuildFuse.<init>(Unknown Source)
    at chemaxon.marvin.modelling.build.BuildSequence$BuildFuse.<init>(Unknown Source)
    at chemaxon.marvin.modelling.build.BuildSequence.<init>(Unknown Source)
    at chemaxon.marvin.modelling.build.FragClean.startCleanOnSingleFragment(Unknown Source)
    at chemaxon.marvin.modelling.interfacing.InvokeFFFBuildCalculation.invoke(Unknown Source)
    at chemaxon.marvin.modelling.interfacing.MultifragMoleculeAdapter.invoke(Unknown Source)
    at chemaxon.marvin.modelling.interfacing.CxnMoleculeAdapter.invoke(Unknown Source)
    at chemaxon.marvin.modelling.interfacing.InvokeCalculation.invoke(Unknown Source)
    at chemaxon.marvin.modules.Clean3D.modfunc_0(Unknown Source)
    at chemaxon.marvin.modules.Clean3D.modfunc(Unknown Source)
    at chemaxon.struc.MoleculeGraph.clean(Unknown Source)
    at chemaxon.struc.Molecule.clean(Unknown Source)
    at chemaxon.marvin.calculations.ConformerPlugin.calculateConformers(Unknown Source)
    at chemaxon.marvin.calculations.ConformerPlugin.run(Unknown Source)
    at chemaxon.marvin.plugin.concurrent.PluginWorkUnit.call(Unknown Source)
    at chemaxon.marvin.plugin.concurrent.ReusablePluginWorkUnit.call(Unknown Source)
    at chemaxon.util.concurrent.marvin.CompositeWorkUnit.call(Unknown Source)
    at chemaxon.util.concurrent.processors.InputOrderedWorkUnitProcessor.process(Unknown Source)
    at chemaxon.util.concurrent.processors.InputOrderedWorkUnitProcessor.processInput(Unknown Source)
    at chemaxon.util.concurrent.processors.WorkUnitWorker.work0(Unknown Source)
    at chemaxon.util.concurrent.processors.WorkUnitWorker.work(Unknown Source)
    at chemaxon.util.concurrent.worker.Worker$1.call(Unknown Source)
    at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:334)
    at java.util.concurrent.FutureTask.run(FutureTask.java:166)
    at chemaxon.util.concurrent.worker.Worker.run(Unknown Source)
    at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
    at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:334)
    at java.util.concurrent.FutureTask.run(FutureTask.java:166)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1110)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:603)
    at java.lang.Thread.run(Thread.java:679)
Clean failed.

Hi, 

 

Thanks for the report, we started to investigate the problem.

Sorry for the late relpy.

 

Regards, 

Peter

Thanks Peter for taking a look at it :)

Dear Supriya,

 

Thanks for the bug report. We will notify You in this topic when this bug is fixed.

It seems that an internal error in the 3D builder caused this failure. As a workaround you can try using our simpler MM based tools (geometry optimization, molecular dynamics) to try to generate valid 3D conformations.

• Simply using "Geometry optimization" on the 2D structures as cleaning method might lead to "tangled" conformations, however for a couple of the input structures from the attachad file it results in good structures


• The tangled structures sometimes can be "straightened" using drastic MD perturbation followed by geometry optimization:
  
  
  
  • For a tangled conformation use tools / conformers / molecular dynamics. Parameters:
    
    
    
    
    • Simulation steps: 1000
    
    
    • Step time: 5fs
    
    
    • Initial temperature 3000K
    
    
    • Start time of display: 0fs
    
    
    • Frame interval: 100fs
    
    
    
    
  
  
  • Using MD with the unrealistic parameters above results in heavily distorted structures
  
  
  • Use "Geometry optimization" on the first "invalid" conformation
  
  
  • This perturbation-optimization approach seems to work on most of the tangled 3D conformations produced by Geometry optimization alone
  
  
  
  

Sorry for the inconvenience; if you have any further questions please feel free to contact us,

 

Regards,

Gabor

Dear Supriya,

 

Using the workaround method described earlier i was able to generate 3D geometry for all of the structures. (Note that it is might better to use 4fs step time).

The structures in 3D are attached in sdf format.

 

regards,

Gabor

Hi Gabor,

Thank you for working on it. However I could not figure out how to do the geometry optimization. Do I need to download a plugin or something? I could not find it under any of the tabs in marvin view.

Moreover, are all of these steps possible to execute in command line as well?

Dear Supriya,

 

Sorry for the late answer

However I could not figure out how to do the 

geometry optimization. Do I need to download a plugin or something? I 

could not find it under any of the tabs in marvin view.

In mview right click on the structure/cell to bring up the context menu and make option "Structure" - "Clean 3D" - "Cleaning method" - "Gradient optimize" be selected, then rigth click again and select "Structure" - "Clean 3D" -"Clean in 3D".

Moreover, are all of these steps possible to execute in command line as well?

You can invoke geometry optimization in molconvert using

molconvert -3:c2o1 sdf [INPUTFILE]

You can invoke MD calculation from cxcalc, however since the workaround described before needs extensive manual supervision i would recommend the GUI method.

 

Regards,

Gabor