R-Group Decomposer

Hi Mikhail,

Many many thanks for all these new nodes recently, I have still not had chance to play with them all yet.

However, I do have a request around an existing node, the R-Group Decomposer.

Is it possible to make it so that when a core is specified and the R-Groups are decomposed into separate columns, that the R-Groups are separated out consistently into the same columns.

For instance, if you have a pyridine core with groups around it, can it be achieved in such a way that the R-Groups are positioned into the same R column for each row. I have numerous examples, where the R-Group splitting seems to randomly change, therefore meaning a common R group in each of the molecules attached to the same position on a pyridine suddenly ends up going from the R-Group #1 column to the R-Group #5 column. I assume this is because the numbering around the Pyridine can be counted from either way round.

One way this might be achieved is by comparing results from the last 5 rows to look for the same or similar group, to help determine which way around the Pyridine to count and therefore to help the R-Group columns be moved into consistently the same column.



Dear Simon,

We are working with RGroup decomposer at the moment. There are several requests (you can find it in the google groups) for the indigo decomposer, so, I think, the KNIME node version will be uploaded soon. I have reproduced the issue with Pyridine. The issue can be expanded on all highly-simmetric scaffold molecules. I will try to resolve it as well.

With best regards,


Hello all,

The bug with RGroup decomposer highly-simmetric scaffold molecules was fixed. The new version will be available in the upcoming release

PS. After an absence we have returned to the KNIME development! I will try to resolve all the issues collected from the forum topics

With best regards,


GGA Software Services LLC.