ArrayIndexOutOfBoundsException when creating mrv

java.lang.ArrayIndexOutOfBoundsException: -1


at chemaxon.marvin.modules.SSSR.fillRingIdx(SSSR.java:1017)


at chemaxon.marvin.modules.SSSR.getRings(SSSR.java:396)


at chemaxon.marvin.modules.SSSR.callback(SSSR.java:1179)


at chemaxon.struc.CGraph.getSSSR(CGraph.java:1189)


at chemaxon.marvin.modules.Parity.initMolecule(Parity.java:1926)


at chemaxon.marvin.modules.Parity.modfunc(Parity.java:190)


at chemaxon.struc.MoleculeGraph.getParity(MoleculeGraph.java:1791)


at chemaxon.marvin.modules.CmlExport.areAtomParitiesWritten(CmlExport.java:375)


at chemaxon.marvin.modules.CmlExport.appendMoleculeGraph(CmlExport.java:348)


at chemaxon.marvin.modules.CmlExport.appendMolecule0(CmlExport.java:266)


at chemaxon.marvin.modules.CmlExport.appendReaction(CmlExport.java:219)


at chemaxon.marvin.modules.CmlExport.appendMolecule(CmlExport.java:169)


at chemaxon.marvin.modules.MrvExport.appendChemicalStruct(MrvExport.java:294)


at chemaxon.marvin.modules.MrvExport.convert0(MrvExport.java:95)


at chemaxon.marvin.modules.MrvExport.convert(MrvExport.java:71)


at chemaxon.struc.Molecule.exportToObject(Molecule.java:1324)


at chemaxon.struc.Molecule.exportToObject(Molecule.java:1289)


at chemaxon.struc.Molecule.exportToFormat(Molecule.java:1138)


at chemaxon.struc.Molecule.toFormat(Molecule.java:1116)





Sorry but I don't have a structure example yet

Thanks, will notice the colleague who responsible for the SSSR code seems could not import the structure.

Could you please send the structure.


We need some information to reproduce and fix the bug.





thank you


Andras

I've added some code to catch this and will try and deploy it to production tomorrow - it is hard to debug these runtime exceptions...

We have also tried to find the bug using the NCI database, but without success.





Andras

I have attached a test code which prints out the molecule number and run sssr detection.


The buggy molecule is right before the exception.





Copy the RingTest.java under the test directory or remove the first line ("package test").








Andras

We are only seeing this error when creating mrvs of reactions - not individual molecules - i am working on getting a structure to you - but this error is causing up big problems.

Hi!





I'm sorry but this is not enough information us to find the bug.


Please send the reaction mrv file or any information how can we reproduce the bug.


Andras

I understand but it is a little bit of a catch 22 in that I can't capture the mrv to reproduce this because I can't generate it - I've captured the molfile and can't reproduce it based on that. We have had to disable the feature that depends on this because it was causing our users so many problems so it is even harder to find now.

We have just found a bug that has maybe connection to this one.


It also appears during reaction generation, but it is reproducible with simple code using jchem API.(It is not possible to generate molecule and get the exception after.) So in that case the code is the only possibility to get the exception.


You may also consider to write a simple code which would generate the exception. As it is possible that you cannot catch the buggy molecule as it is generated somewhere in the code.

The related problem was the following:


It is not supported to embed different types of molecules in any way.


What does this mean:


If you want a reaction with Rgroups you should set up an RgMolecule and set the root of the RgMolecule to RxnMolecule.





The following for example is not supported:


Setting up an RxnMolecule which contains RgMolecule as a reagent and Molecule as product.





I hope this helps.


Andras

Dear Andras,





I have the same problem, but with standardizer. Here I am attaching a molecule which will reproduce the error in the code posted by you. I just added the following line to it after line no 29





System.out.println(m.toFormat("smiles"));





I come across this problem in any programmatic attempt to access this molecule object, not only in standardizer.





Thanks and regards,





Ravindra





*************************************************


jchem.version=5.1.0


jchem.veruln=5_1_0


jchem.vernum=5.1.0


jchem.date=2008.08.07


jchem.table.version=50100


marvin.package=marvin-all-5_1_0.zip





os=MS Windows Server 2003 RC2 SP2


java version "1.6.0_10"


Java(TM) SE Runtime Environment (build 1.6.0_10-b33)


Java HotSpot(TM) Client VM (build 11.0-b15, mixed mode, sharing)


*************************************************

Hi!





Andras is on holiday now, but we will check this soon.





Best Regards,


Peter

Dear Ravindra,





We're sorry for the late answer.


Thank you for reporting, this is indeed a bug in SMILES export.


We will notify you as soon as the correction is out.





Best Regards,


Erika

Dear Erika,


Thanks for the reply. I can send you couple of more sample MOL files for this error, if that helps.





Best Regards,





Ravindra

More sample files would help us a lot.





Thanks,


Erika.

Hi Ravindra,





I've come across this topic, and tried out your test file (tes.mol). To be honest, I don't understand why did you create a superatom from the same structure three times, in addition, why did you create one of these superatoms on the top of another. Anyway, it seems that Marvin incorrectly imports this MOLFile, because after expanding the groups, the original structure is duplicated. We have to investigate this issue first.





To reproduce the SMILES export problem, could you please send us other test files?





Thank you and best regards,





Akos