R-Group decomposition and Fragment


I'm not really a chemoinformaticien but more a chemist. I'm used to write on forum, but I have a 'true' limitation...

I'm trying to do a R-Group decomposition with Indigo node for populate a table with both R-Group and Fragment. This is essential to build the spotfire guide I want.

If I may. A couple of improvement of the R-Group node (already really useful) would be:

* consider hydrogen as a group

* be able to introduce the Rs on the query molecules

* get R-Group and Fragment in the output table.

I did the R-Group decomposition but I'm not able to get the corresponding fragment... I tried with RDkit reactions, but it's quite tricky... and to be honnest, I don't have any idea for that.

any one tried to do something similar?

many thanks for your help. and thanks for Indigo nodes, very useful already.






I uploaded a workflow into the public server called complex SAR analysis, if I recall correctly.

that uses the indigo R group decomposition node. Does this help with your problems.


You can simply get the Hydrogen atom as a substituent by using a Missing Value node, and use Hydrogen as the replacement text.




Many thanks for the trick helping me to add H in the table. it works well.

For the fragment, your nice protocol 'complexSARanalysis' doesn't help me really. (I had this protocol since a while, it helped me a lot for building loops, for instance).

I'm trying to get a table like the following after a R Group decomposition:

and I have almost 14 scaffolds with 6 R Group each...


many thanks again !




I am not fully understanding the issue.

In the output from the Indigo Decomposition node, you get all the R groups, and the scaffold column too. Isnt the scaffold column the same as the fragment column you mention ?


Good point Simon. exemple was not precise enough. Difficulties arise when more than 1 RGroup is considered.


I meant the following:

 You have this kind of information by using MMP nodes from Erl Wood, but that take me off phylosophy of RGroup decomposition. I thought to do that any way and cross-mach both protocols to get a table of MMP filtered by RGroup I want, single raw by compound.

Also, make a library of reaction to 'rebuild' fragments from Rgroups and Scaffold or  even, make a'soustraction' of molecules by RGroup to get Fragment.

I tried all of these approaches on model, but they're not really convenient/practical... complicate a lot my life...

Any idea more than welcome,


many thanks !




To create fragments for use in reactions to rebuild molecules, use the Atom Replacer node on the output molecules from the R Group Decomposition node. Choose to replace the attachment points with an At atom.

Then convert your fragments back to SDF or Smiles and then convert to RDKit. You can then use these fragments with the At group to rebuild molecules using either the one component reaction or two component reaction nodes.

to define the the reaction in a one component reaction node, use the Marvin Sketcher node and draw an At atom with a methyl group coming off, and on the right side of the reaction arrow draw the rest of the molecule with an At atom coming off the position where you want your groups/fragments to go. Finally you need to map the reaction atoms, so using the arrow icon, drag from the methyl group next to the At atom on the left side to the atom next to the At group on the right side. This should put a little number 1 on both atoms.

hope this is helpful for you.


This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.