Enantiomer enumeration in RDKit reaction nodes

Hi all,

A reaction I have encoded in a pair of Two component reaction nodes (formation of a 4H-benzopyran from an aromatic aldehyde, an acetonitrile and a phenol) that creates a new stereocentre.  The node gives no option to enumerate both enantiomers for this reaction.  I tried  drawing the reaction in MarvinSketch with explicitly defined stereocentres, and using chiral SMARTS [C@@H:2]. No joy.

Would it be possible to add this feature the the reaction node?  Or can someone suggest a workaround? (I tried the Indigo enumerate isomers node but that didnt work unfortunately).

Alastair

The indigo enumerate isomers node should do what you need.  I cant remember which way around it is, either you need a flat bond or a wedge bond for the enumeration to work.

Simon.

This is possible, though it is perhaps not as convenient as it might be. The attached workflow shows how.

If you express the reaction using SMARTS and have a C in the reactants that maps to a C in the products with a stereochem spec (i.e. @ or @@), then you'll end up with a stereocenter in the final molecule. You need to have two reaction nodes to get both stereoisomers.

There may be some way to do this with RXNs (i.e. so that you can draw your reaction in Marvin Sketch), but I didn't test that.

Note: there seems to be a bug in the RDKit rendering of the output molecules so you don't see wedged bonds in the depiction. If you switch to a string view of the molecules in the product column, you will see the stereocenters in the products.

Does this do what you're looking for?

As an aside: it's probably worth thinking about adding an "expand stereo-isomers" node. This functionality does not exist in the RDKit but it's relatively easy to implement.

-greg

 

Hi Greg, this does what I want thanks.  The inconvenience of having parallel reaction schemes isnt too bad....and perhaps easier than coding for reactions with multiple stereocenters being formed.

Alastair