Chemaxon issues with using MolImporter in existing nodes?

Groups Cheminformatics
1

I'm not sure what the issue is here, I'm posting here rather than at Chemaxon because it seems to specifically affect the library within KNIME rather than Marvin itself.

 

I use the Chemaxon functions to output a series of structures in SMILES format, such as this 

O=C(Nc1ccncc1)c1ncnc1C(=O)OC1CCCCC1

Now this is actually invalid, because of the ambiguity around the five memebred ring, but I can't seem to find a combination of functions that will let me fix this.

 

The reason I bring it up here is that if you were to import this string into MarvinSketch a valid structure is produced, and indeed Chemaxon renders it, but copying the SMILES from that gives O=C(Nc1ccncc1)c1[nH]cnc1C(=O)OC1CCCCC1, which is perfectly valid.

Is there a snippet of code or something that can be done to clean this up within a node, I've tried the various options of MolImporter to copy the structure into a new Molecule instance, but the invalid form remains.

I would have thought that one of the dearomaize/aromatize combinations would be enough to clear it, but it seems to get 'stuck' in an ambigous state.

2

Dear James_Wallace,

The reason for the difference is that the Marvin Sketch GUI uses non-default valence check options, which fixes such ambiguous N atoms, and adds an implicit H if needed.
We normally do not do this "fix" automatically (e.g. "molconvert" does not modify the structure): hidden, automatic structure modifications should be avoided in general (of course drawing is different).
For explicit structure checking and modification I would suggest Structure Checker, using the "Valence Error Checker" it can fix such structures.

https://docs.chemaxon.com/display/docs/Valence+Error

Best regards,
Taka

3