Hi Rocco, On Tue, Jan 5, 2016 at 12:26 AM, Rocco Moretti <[email protected]> wrote:
> > It doesn't look like user-set properties are copied through with > explicitly mapped atoms in chemical reactions, although implicitly copied > atoms do have their properties transfered. (See example session below.) > > Is this the intended behavior? > It's intended in the sense of "yes, that's how it was written". Having said that, I can't come up at the moment with any compelling reason not to change it though. > If I were to want all mappable atoms (implicit & explicit) to have > particular properties transfered across the reaction, how best would I go > about doing so (specifically on the C++ level)? > After a quick scan through the code, I believe that the changes would be made in this function: https://github.com/rdkit/rdkit/blob/master/Code/GraphMol/ChemReactions/ReactionRunner.cpp#L388 I'm happy to add the required bits of code, but I want to spend a bit more time thinking about the consequences. Feel free to go ahead and create the github issue requesting the feature though. :-) -greg > Thanks, > -Rocco > > -- > > from rdkit import Chem > from rdkit.Chem import AllChem > > def print_arb_prop(mollist): > for m in range(len(mollist)): > for a in range(mollist[m].GetNumAtoms()): > atom = mollist[m].GetAtomWithIdx(a) > if atom.HasProp("arb_prop"): > print m, a, atom.GetSymbol(), atom.GetIntProp("arb_prop") > > mol = Chem.MolFromSmiles("ClCNOCBr") > for a in range(mol.GetNumAtoms()): > mol.GetAtomWithIdx(a).SetIntProp("arb_prop", a*-2 ) > > print_arb_prop( [mol] ) > # 0 0 Cl 0 > # 0 1 C -2 > # 0 2 N -4 > # 0 3 O -6 > # 0 4 C -8 > # 0 5 Br -10 > > rxn1 = AllChem.ReactionFromSmarts('[N:3][O:4]>>[N:3].[O:4]') > print_arb_prop( rxn1.RunReactants([mol])[0] ) > # 0 1 C -2 > # 0 2 Cl 0 > # 1 1 C -8 > # 1 2 Br -10 > > rxn2 = > AllChem.ReactionFromSmarts('[Cl:1][C:2][N:3][O:4][C:5]>>[Cl:1][C:2][N:3].[O:4][C:5]') > print_arb_prop( rxn2.RunReactants([mol])[0] ) > # 1 2 Br -10 > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > Rdkit-discuss mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/rdkit-discuss > >
------------------------------------------------------------------------------
_______________________________________________ Rdkit-discuss mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
