You are absolutely correct. There's a typo in that function. I'll fix it.
-greg
On Mon, Jun 20, 2016 at 5:43 PM, Luca Fenu - Network <
fenu_l...@network.lilly.com> wrote:
> Thanks Greg,
>
>
>
> that works great. I’ve reconverted the isotopic smarts to smiles with the
> function suggested in (
> http://rdkit.org/docs/Cookbook.html#using-custom-mcs-atom-types):
>
>
>
> *def* getMCSSmiles(mol,labelledMol,mcs):
>
> mcsp = Chem.MolFromSmarts(mcs.smartsString)
>
> match = labelledMol.GetSubstructMatch(mcsp)
>
> *return* Chem.MolFragmentToSmiles(ms[0],atomsToUse=match,
>
> isomericSmiles=*True*,
>
> canonical=*False*)
>
>
>
> However, I think there may be a typo, and it should be:
>
>
>
> *def* getMCSSmiles(mol,labelledMol,mcs):
>
> mcsp = Chem.MolFromSmarts(mcs.smartsString)
>
> match = labelledMol.GetSubstructMatch(mcsp)
>
> *return* Chem.MolFragmentToSmiles(*mol*,atomsToUse=match,
>
> isomericSmiles=*True*,
>
> canonical=*False*)
>
>
>
> I’ll see if I can use the custom atom types to get rid of more unwanted
> matches, such as those you highlighted in my previous email.
>
>
>
> Thanks again,
>
>
>
> *Luca Fenu – Computational Chemist*
>
> Computational Chemist, CADD Group, Chemistry
> Eli Lilly and Company Limited
> Erl Wood Manor, Sunninghill Road, Windlesham, Surrey, GU20 6PH
> 012.764.83354 (office) | 07858.701.646 (mobile)
> fenu_l...@network.lilly.com <surname_n...@network.lilly.com> |
> www.lilly.co.uk
>
>
>
> [image: cid:image005.jpg@01D0842C.DF674960]
>
>
>
> *From:* Greg Landrum [mailto:greg.land...@gmail.com]
> *Sent:* 18 June 2016 05:16
> *To:* Luca Fenu - Network
> *Cc:* rdkit-discuss@lists.sourceforge.net
> *Subject:* Re: [Rdkit-discuss] MCS on fused ring systems
>
>
>
> Hi Luca,
>
>
>
> On Fri, Jun 17, 2016 at 5:28 PM, Luca Fenu - Network <
> fenu_l...@network.lilly.com> wrote:
>
>
>
> I’m back again with what seems to me like puzzling behaviour from the
> rdfMCS.findMCS function when dealing with fused ring systems:
>
>
>
> > mol3RA = Chem.MolFromSmiles('c1ccc2cc3ccccc3cc2c1')
>
> > mol3RB = Chem.MolFromSmiles('c1ccc2c(c1)ccc1ccccc21')
>
> > rdFMCS.FindMCS([mol3RA, mol3RB], matchValences=True,
>
> matchChiralTag=True,
>
> ringMatchesRingOnly=True,
>
> completeRingsOnly=True # False will return the same result
>
> ).smartsString
>
> '[#6]-[#6]1:[#6]:[#6]:[#6]:[#6]:[#6]:1'
>
> Or, in pictures:
>
> The MCS shown above is the same whether or not I turn on the
> ‘completeRingsOnly flag.
>
> So, the atoms are all there, but the MCS ring can’t be closed because the
> attachment point doesn’t match.
>
> Since we’re requesting completeRingsOnly, wouldn’t make sense for the MCS
> be reduced to just a naphthalene?
>
> What’s the rationale behind the current implementation? Is there a way to
> force it to drop the open circle?
>
>
>
> That's a bug in the current implementation. I'll take a look at it and see
> if I can track it down.
>
>
>
>
>
> And another, related, question: is there a way to make the MCS computation
> even more restrictive,
>
> only returning a ring as part of the MCS *if the whole fused system is
> matched*, not just a portion (albeit a complete ring by itself?). In the
> example described above, there would be no MCS.
>
>
>
> I’d be fine with some post-processing to the same aim. Another example:
>
>
>
>
>
> That's an interesting question. I think you can get at least part of what
> you're looking for using the "atom labeling" trick from here:
> http://rdkit.org/docs/Cookbook.html#using-custom-mcs-atom-types
>
> Here's a pass at the first step of that:
>
>
>
> In [70]: def label(a):
>
> return 100*a.GetOwningMol().GetRingInfo().NumAtomRings(a.GetIdx()) +
> a.GetAtomicNum()
>
> ....:
>
>
>
> In [71]: bz = Chem.MolFromSmiles('c1ccccc1')
>
>
>
> In [72]: napth = Chem.MolFromSmiles('c1ccc2ccccc2c1')
>
>
>
> In [73]: rdFMCS.FindMCS([bz,napth]).smartsString
>
> Out[73]: '[#6]1:[#6]:[#6]:[#6]:[#6]:[#6]:1'
>
>
>
> In [74]: nms = [Chem.Mol(x) for x in (bz,napth)]
>
>
>
> In [75]: for nm in nms:
>
> ....: for at in nm.GetAtoms():
>
> ....: at.SetIsotope(label(at))
>
> ....:
>
>
>
> In [76]: mcs =
> rdFMCS.FindMCS(nms,atomCompare=rdFMCS.AtomCompare.CompareIsotopes)
>
>
>
> In [77]: mcs.smartsString
>
> Out[77]: '[106*](:[106*]:[106*]):[106*]'
>
>
>
> In [78]: mcs =
> rdFMCS.FindMCS(nms,atomCompare=rdFMCS.AtomCompare.CompareIsotopes,completeRingsOnly=True)
>
>
>
> In [79]: mcs.smartsString
>
> Out[79]: ''
>
>
>
> Does that help?
>
>
>
> -greg
>
>
>
>
>
------------------------------------------------------------------------------
Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San
Francisco, CA to explore cutting-edge tech and listen to tech luminaries
present their vision of the future. This family event has something for
everyone, including kids. Get more information and register today.
http://sdm.link/attshape
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
