exception generated when predicting pKa with user trained -

Hi, I got this exception when predicting pKa trained with our own data. If I don't use the trained model, the prediction is fine.

 

"java.lang.ArrayIndexOutOfBoundsException: -1\n    at chemaxon.calculations.pka.IntrinsicIonization.setSemiAraInc(IntrinsicIonization.java:9386)\n    at chemaxon.calculations.pka.IntrinsicIonization.setHBondTable(IntrinsicIonization.java:5428)\n    at chemaxon.calculations.pka.Ionizer.getMicropKa(Ionizer.java:1217)\n    at chemaxon.calculations.pka.Ionizer.calcSpeciesRatio(Ionizer.java:3245)\n    at chemaxon.calculations.pka.Ionizer.calcSpeciesOrder(Ionizer.java:3417)\n    at chemaxon.calculations.pka.Ionizer.calcMsDistribution(Ionizer.java:3651)\n    at chemaxon.calculations.pka.Ionizer.calculatepKa(Ionizer.java:3975)\n    at chemaxon.calculations.pka.Ionizer.calculatepKa(Ionizer.java:3894)\nat chemaxon.marvin.calculations.pKaPlugin.run(pKaPlugin.java:767)\n

 

any idea?

 

fengjun wrote:

Hi, I got this exception when predicting pKa trained with our own data. If I don't use the trained model, the prediction is fine.   "java.lang.ArrayIndexOutOfBoundsException: -1\n    at chemaxon.calculations.pka.IntrinsicIonization.setSemiAraInc(IntrinsicIonization.java:9386)\n    at chemaxon.calculations.pka.IntrinsicIonization.setHBondTable(IntrinsicIonization.java:5428)\n    at chemaxon.calculations.pka.Ionizer.getMicropKa(Ionizer.java:1217)\n    at chemaxon.calculations.pka.Ionizer.calcSpeciesRatio(Ionizer.java:3245)\n    at chemaxon.calculations.pka.Ionizer.calcSpeciesOrder(Ionizer.java:3417)\n    at chemaxon.calculations.pka.Ionizer.calcMsDistribution(Ionizer.java:3651)\n    at chemaxon.calculations.pka.Ionizer.calculatepKa(Ionizer.java:3975)\n    at chemaxon.calculations.pka.Ionizer.calculatepKa(Ionizer.java:3894)\nat chemaxon.marvin.calculations.pKaPlugin.run(pKaPlugin.java:767)\n   any idea?

Also, this only happens in API call or use cxcalc command line version. If I just paste the molecule within marvin sketch, and do calculation within it, it works fine.

 

Hi, 

We would need some molecules that failed to reproduce the error message. Can you send us any? 

We also need to know which version of cxcalc you use.

Thank you very much, 

Daniel

dszisz wrote:

Hi,  We would need some molecules that failed to reproduce the error message. Can you send us any?  We also need to know which version of cxcalc you use. Thank you very much,  Daniel

Attached a fake molecule, and the trained model. the version is 15.12.14.0

dszisz wrote:

Hi,  We would need some molecules that failed to reproduce the error message. Can you send us any?  We also need to know which version of cxcalc you use. Thank you very much,  Daniel

Any update? is the problem reproducible on your end? Is the problem already fixed in the newest version? thanks.

Hi, 

We have managed to reproduce the problem, but we need time to provide fix for that. The problem seems to exist in the newer versions as well. We will fix it ASAP. 

Daniel

dszisz wrote:

Hi,  We have managed to reproduce the problem, but we need time to provide fix for that. The problem seems to exist in the newer versions as well. We will fix it ASAP.  Daniel

Hi, any update on this? thanks,

Hi, 

Our colleague who deals with this issue is on holiday now. We will need some time to sort it out. 

Thank you for your patience. 

Daniel