Substructure matching


I have 2 scenarios

1) I've been running a substructure search where I wanted to look for a substructure which had a specific charge as I was checking for errors (a 5-mem ring that contained N=N+C-). The CDK node seemed to wipe out this charge and didn't find anything that matched that conformation, even though I know there was at least one matching compound in there.


2) I was also looking for a set of ribose compounds and filtered on a basic ribose substructure. I then wanted to filter further and added another substructure node where I added specific stereochemistry to this ribose and it just ignored the stereochemistry and passed everything through.

So my question is, do you not have an option for exact matching that takes into account charges and stereochemistry?





the new substructure search takes stereo- and charge information into account. I have updated the nightly build to that effect.

It matches stereoinformation if defined, otherwise the search does not discriminate between bond types. Matching of formal atom charges is optional and can be turned on.

I hope that takes care of both your scenarios. I would be grateful if you could check out the changes and let me know if the search now works for you. I have tested various charge/stereo combinations with a small set of compunds.



Hi Stephan

I've run the updates on Knime and now stereochemistry isn't retained in the drawing window.

It also keeps falling over and won't process my file. 

It was passing my file without incident on friday.






This is the error text that I get


ERROR     Substructure Search                Execute failed: Invalid result. Execution failed, reason: data at output 0 is null.

Hi Louisa,

yes, the changes didn't go through. My apologies. I will push the changes again.