error in the logp calculation

User 151fddb1df

27-06-2011 11:21:16

Hi,


I tried to use chemaxon.marvin.calculations.logPPlugin to caculate the logP descriptor of a molecule but it did not work for some molecule. I tried with marvinSketch applet on your website with no luck.


Here is the error :


ERROR:
Cannot run plugin: logP
-1

STACK TRACE:
java.lang.ArrayIndexOutOfBoundsException: -1
    at chemaxon.calculations.pKa.t(Unknown Source)
    at chemaxon.calculations.pKa.aa(Unknown Source)
    at chemaxon.calculations.pKa.E(Unknown Source)
    at chemaxon.calculations.pKa.C(Unknown Source)
    at chemaxon.calculations.pKa.getAcidicpKa(Unknown Source)
    at chemaxon.calculations.Ionizer.h(Unknown Source)
    at chemaxon.calculations.Ionizer.e(Unknown Source)
    at chemaxon.calculations.Ionizer.initMsCalculation(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.a(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.predict(Unknown Source)
    at chemaxon.marvin.calculations.logPPlugin.run(Unknown Source)
    at chemaxon.marvin.plugin.CalculatorPluginManager.a(Unknown Source)
    at chemaxon.marvin.plugin.CalculatorPluginManager.a(Unknown Source)
    at chemaxon.marvin.plugin.CalculatorPluginManager.a(Unknown Source)
    at chemaxon.marvin.plugin.CalculatorPluginManager.callback(Unknown Source)
    at chemaxon.marvin.plugin.PluginManager.callback(Unknown Source)
    at chemaxon.marvin.common.swing.MolPanel$5.run(Unknown Source)


 


Do you hava any idea ? it did not seem to come from the code of my application and as for the molecule, the previous version of JChem and Marvin Sketch did not mind it and the logP was calculated.


Thanks for your help.


L.

User 851ac690a0

27-06-2011 12:10:36

Hi,


Please send a "test" molecule so that we can try to reproduce your error message.


Thanks.


Jozsi

User 151fddb1df

27-06-2011 12:17:14

Hi,


Thanks for your answer.


do you have an email adress to wich I can send my file please ?



L.

User 851ac690a0

27-06-2011 12:48:39

Hi,


Thank you for the test molecule.  


We fixed a bug in the pKa calculator.


The fixed version will be available soon.  In the 5.5.1 version.


Jozsi

User 151fddb1df

27-06-2011 13:04:35

Hi,


thanks a lot.


Any idea when the next release will occur ? in a few days or weeks ?


 


L.

User 851ac690a0

27-06-2011 15:43:44

Hi,


One of my collegaue is going to answer for this question.


Jozsi

ChemAxon e08c317633

29-06-2011 09:33:08

5.5.1 will be released in first half of July 2011.


Zsolt

User dfeb81947d

04-10-2011 13:16:05

Dear Support,


I'm using JChem 5.5.1 under windows XP.


I also have a problem with logP calculation.


In the SDF file attached to this message, the three first molecules are peptides, and the logP calculation failed with the Stack Trace


java.lang.NullPointerException
    at chemaxon.calculations.Ionizer.setDegIndex(Unknown Source)
    at chemaxon.calculations.Ionizer.assignAtom(Unknown Source)
    at chemaxon.calculations.Ionizer.calcpKa(Unknown Source)
    at chemaxon.calculations.Ionizer.calculatepKa(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.createIonizer(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.predict(Unknown Source)
    at chemaxon.marvin.calculations.logPPlugin.run(Unknown Source)

I can guess why (regarding the monstrous structure)... so you don't need to feel concern with it, it was just for your information.


The other structures fails with StackTrace as follow


java.lang.ArrayIndexOutOfBoundsException: 2
    at chemaxon.calculations.pKa.getAvarageAliphaticRingCharge(Unknown Source)
    at chemaxon.calculations.pKa.getOHAcidicpKa(Unknown Source)
    at chemaxon.calculations.pKa.getAlcoholicAcidicpKa(Unknown Source)
    at chemaxon.calculations.pKa.getApKa(Unknown Source)
    at chemaxon.calculations.pKa.getAcidicpKa(Unknown Source)
    at chemaxon.calculations.Ionizer.initMicropk0(Unknown Source)
    at chemaxon.calculations.Ionizer.calcStartValues(Unknown Source)
    at chemaxon.calculations.Ionizer.initMsCalculation(Unknown Source)
    at chemaxon.calculations.Ionizer.isMacropKaCalcRequiredBylogP(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.isMacropKaCalcRequiredBylogP(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.createIonizer(Unknown Source)
    at chemaxon.calculations.training.logp.LogPPredictorImpl.predict(Unknown Source)
    at chemaxon.marvin.calculations.logPPlugin.run(Unknown Source)

(calculation fails on pKa calculation too)


But... for those molecule, logP calculation succeed with JChem 5.2.6... I don't have JChem 5.6.0.1 installed yet, so haven't tried with it (sorry -_-; )


I hope this help


Best Regards,


Jacques

ChemAxon afdac7b783

05-10-2011 19:15:41

Dear Jacques,


We are investigating the issue and we will respond soon.

Kind regards,


Viktoria

ChemAxon afdac7b783

07-10-2011 06:50:33

Dear
Jacques,


I could
reproduce your first exception in v. 5.5.1 and v. 5.6.0.1 as well. Your guess
was right; the problem is found and the fix will be out in version 5.8.


Thank
you for reporting this issue.


Regarding
your next issue: I could reproduce the second problem only in v. 5.5.1; this exception
does not take place in v. 5.6.0.0 or in v. 5.6.0.1.


Until the
new version is out, I would recommend using v. 5.6.0.1 to calculate logP (or
pKa) data.
cxcalc with option -g (continue with next molecule on error) will
skip macromolecules but will provide results for the other molecules in your
file.


cxcalc -g logp molecule_error_logP.sdf


Best
regards,


Viktoria