Re: [Rdkit-discuss] Do we have an exact implementation of Bemis-Murcko scaffolds in rdkit?

[Greg L. <gre...@gm...>]

I'm not sure why you'd want to reimplement something that's already there,
but if this works better for you...

the easiest way to get a single function you could call would be to do
something like:

In [18]: def MolToGenericScaffold(mol):
    ...:     return
MurckoScaffold.MakeScaffoldGeneric(MurckoScaffold.GetScaffoldForMol(mol))
    ...:
In [19]:
Chem.MolToSmiles(MolToGenericScaffold(Chem.MolFromSmiles('CCc1ccc(O)cc1C(=O)C1CC1')))
Out[19]: 'CC(C1CCCCC1)C1CC1'


-greg


On Tue, Apr 27, 2021 at 4:32 AM Francois Berenger <ml...@li...> wrote:

> On 27/04/2021 10:12, Francois Berenger wrote:
> > On 26/04/2021 23:35, Greg Landrum wrote:
> >> Hi Francois,
> >>
> >> The implementation which is there does, I believe, the right thing.
> >> However... first you need to find the Murcko Scaffold, then you can
> >> convert that scaffold to the generic form:
> >>
> >>> In [5]: m = Chem.MolFromSmiles('CCc1ccc(O)cc1C(=O)C1CC1')
> >>> In [6]: scaff = MurckoScaffold.GetScaffoldForMol(m)
> >>> In [7]: Chem.MolToSmiles(scaff)
> >>> Out[7]: 'O=C(c1ccccc1)C1CC1'
> >>> In [8]: framework = MurckoScaffold.MakeScaffoldGeneric(scaff)
> >>> In [9]: print(Chem.MolToSmiles(framework))
> >>> CC(C1CCCCC1)C1CC1
> >
> > Ok, maybe this two steps process is a little bit better, but still
> > not exactly what I would expect in some cases.
> >
> > I'll say if I program something which I prefer.
>
> Hello,
>
> I end up with this:
> ---
> def find_terminal_atoms(mol):
>      res = []
>      for a in mol.GetAtoms():
>          if len(a.GetBonds()) == 1:
>              res.append(a)
>      return res
>
> # Bemis, G. W., & Murcko, M. A. (1996).
> # "The properties of known drugs. 1. Molecular frameworks."
> # Journal of medicinal chemistry, 39(15), 2887-2893.
> def BemisMurckoFramework(mol):
>      # keep only Heavy Atoms (HA)
>      only_HA = rdkit.Chem.rdmolops.RemoveHs(mol)
>      # switch all HA to Carbon
>      rw_mol = Chem.RWMol(only_HA)
>      for i in range(rw_mol.GetNumAtoms()):
>          rw_mol.ReplaceAtom(i, Chem.Atom(6))
>      # switch all non single bonds to single
>      non_single_bonds = []
>      for b in rw_mol.GetBonds():
>          if b.GetBondType() != Chem.BondType.SINGLE:
>              non_single_bonds.append(b)
>      for b in non_single_bonds:
>          j = b.GetBeginAtomIdx()
>          k = b.GetEndAtomIdx()
>          rw_mol.RemoveBond(j, k)
>          rw_mol.AddBond(j, k, Chem.BondType.SINGLE)
>      # as long as there are terminal atoms, remove them
>      terminal_atoms = find_terminal_atoms(rw_mol)
>      while terminal_atoms != []:
>          for a in terminal_atoms:
>              for b in a.GetBonds():
>                  rw_mol.RemoveBond(b.GetBeginAtomIdx(),
> b.GetEndAtomIdx())
>              rw_mol.RemoveAtom(a.GetIdx())
>          terminal_atoms = find_terminal_atoms(rw_mol)
>      return rw_mol.GetMol()
> ---
>
> I don't claim this is very efficient Python code. I am not very good at
> snake charming.
>
> Regards,
> F.
>
> >> Best,
> >> -greg
> >>
> >> On Mon, Apr 26, 2021 at 11:15 AM Francois Berenger <ml...@li...>
> >> wrote:
> >>
> >>> Hello,
> >>>
> >>> I am trying MurckoScaffold.MakeScaffoldGeneric(mol),
> >>> but this keeps the side chains.
> >>>
> >>> While my understanding of BM scaffolds is that only rings
> >>> and ring linkers should be kept.
> >>>
> >>> The fact that the rdkit implementation keeps the
> >>> side chains makes Murcko scaffolds a much less powerful filter
> >>> to enforce molecular diversity.
> >>>
> >>> And I don't even see any option to force the standard/vanilla
> >>> behavior.
> >>> Or, am I missing something?
> >>>
> >>> Regards,
> >>> F.
> >>>
> >>> _______________________________________________
> >>> Rdkit-discuss mailing list
> >>> Rdk...@li...
> >>> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
> >
> >
> > _______________________________________________
> > Rdkit-discuss mailing list
> > Rdk...@li...
> > https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>