Invalid "extended" SMILES from chemical in ChemAxon table

Hi,

We're hitting an issue with ChemAxon searching, where a chemical has been successfully added to the chemicals table (by the UpdateHandler), but the chemical appears to have invalid features which cause exceptions in duplicate filtering.

The extended SMILES in question is:

 

CCC(C)[H]C(=O)C(=O)C([H]C(=O)[C@@H]1[C@H]2CCC[C@H]2CN1C(=O)[C@@H](NC(=O)C([H]C(=O)c1cnccn1)C(C)C)C(C)(C)C)CC(F)F |r,w:9.46,$,;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;$| 

 

I pulled this directly from the chemicals table, and attempted to import it using:

 

MolImporter.importMol(badsmi);

This generates:

Exception in thread "main" chemaxon.formats.MolFormatException: ;: unknown ChemAxon Extended SMILES feature

at chemaxon.marvin.io.formats.smiles.CxsmilesImport.readExtendedFeatures(CxsmilesImport.java:916)

at chemaxon.marvin.io.formats.smiles.SmilesImport.readMol(SmilesImport.java:569)

at chemaxon.marvin.io.formats.smiles.CxsmilesImport.readMol(CxsmilesImport.java:63)

 

This isn't a massive problem for us, but it would be useful to know what the problem is.

J

 

 

Hi James, 

It seems we can't export the exported cxsmiles if it contains an atom alias or atom label which is a "," character.

If you remove that label the IO works fine.

We started to fix the bug, we will get back to you when it's done.

Thanks for the report.

Best regards,

Peter

Hi James, 

 

We managed to fix the bug, from 5.10 we can write and read back "," atom aliases to cxsmiles. 

 

Regards,

Peter