Dear all, An interesting question came up in github ( https://github.com/rdkit/rdkit/issues/2036): The molecule CNC1CCC(c2ccc(Cl)c(Cl)c2)c2ccccc12 [image: image.png] is not split by the BRICS code even though you might expect the bonds between the tetralin ring and both the 6 ring and the amine to be broken.
As explained in my comment in the bug report, the reason nothing happens is because the atom environment that defines the aliphatic-ring carbons that would be involved (L15 in the BRICS paper) - [C;$(C(-;@C)-;@C) - requires both neighbors of the central atom to be aliphatic. We could fix the problem by requiring only one aliphatic neighbor - [C;$(C(-;@C)-;@[#6]). It seems like this change makes sense, but before I do it, I want to make sure no one here sees any fundamental problems with it. Best, -greg
_______________________________________________ Rdkit-discuss mailing list Rdkit-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
