On Mon, Sep 11, 2017 at 5:55 PM, James T. Metz <jamestm...@aol.com> wrote:

> Greg,
>
>     I need to be able to use SMARTS patterns to identify substructures in
> molecules
> that can be aromatic, and I need to be able to handle cases where there
> can be
> differences in the way that the molecule was entered or drawn by a user.
>

That particular problem is a big part of the reason that we tend to use the
aromatic representation of things.


>     For example, consider the following alkenyl-substituted pyridine, there
> are two possible Kekule structures
>
>     m1 = 'C=CC1=NC=CC=C1'
>     m2 = 'C=CC1N=CC=CC1'
>

Fixing what I assume is a typo for m2, I can do the following:

In [11]: m1 = Chem.MolFromSmiles('C=CC1=NC=CC=C1')

In [12]: m2 = Chem.MolFromSmiles('C=CC1N=CC=CC=1')

In [13]: q1 = Chem.MolFromSmarts('cccc')

In [14]: q2 = Chem.MolFromSmarts('cccn')

In [15]: list(m1.GetSubstructMatch(q1))
Out[15]: [2, 7, 6, 5]

In [16]: list(m1.GetSubstructMatch(q2))
Out[16]: [6, 5, 4, 3]

In [17]: list(m2.GetSubstructMatch(q1))
Out[17]: [2, 7, 6, 5]

In [18]: list(m2.GetSubstructMatch(q2))
Out[18]: [6, 5, 4, 3]


Those particular queries were going for the aromatic species and will only
match inside the ring, but if you want to be more generic you could tune
your queries like this:

In [28]: q3 =
Chem.MolFromSmarts('[#6;$([#6]=,:[*])]-,=,:[#6;$([#6]=,:[*])]-,=,:[#6;$([#6]=,:[*])]-,=,:[#6;$([#6]-=,:[*])]')

In [29]: q4 =
Chem.MolFromSmarts('[#6;$([#6]=,:[*])]-,=,:[#6;$([#6]=,:[*])]-,=,:[#6;$([#6]=,:[*])]-,=,:[#7;$([#7]-=,:[*])]')

In [30]: list(m1.GetSubstructMatch(q3))
Out[30]: [0, 1, 2, 7]

In [31]: list(m1.GetSubstructMatch(q4))
Out[31]: [0, 1, 2, 3]

In [32]: list(m2.GetSubstructMatch(q3))
Out[32]: [0, 1, 2, 7]

In [33]: list(m2.GetSubstructMatch(q4))
Out[33]: [0, 1, 2, 3]

If you aren't familiar with recursive SMARTS, this construct:
"[#6;$([#6]=,:[*])]" means "a carbon that has either a double bond or an
aromatic bond to another atom".  So you can interpret q3 as "four carbons
that each have either a double or aromatic bond and that are connected to
each other by single, double, or aromatic bonds".

Is this starting to approximate what you're looking for?
-greg




    Now consider two SMARTS
>
>     pattern1 = '[C]=[C]-[C]={C]
>     pattern2 = '[C]=[C]-[C]=[N]'
>
>     I need to be able to detect the existence of each pattern in the
> molecule
>
>     If m1 is the only available generated Kekule structure, then pattern2
> will be recognized.
>     If m2 is the only available generated Kekule  structure, then pattern1
> will be recognized.
>
>     Hence, I am getting different answers for the same input molecule just
> because
> it was drawn in different Kekule structures.
>
>     Regards,
>     Jim Metz
>
>
>
>
>
> -----Original Message-----
> From: Greg Landrum <greg.land...@gmail.com>
> To: James T. Metz <jamestm...@aol.com>
> Cc: RDKit Discuss <rdkit-discuss@lists.sourceforge.net>
> Sent: Mon, Sep 11, 2017 10:31 am
> Subject: Re: [Rdkit-discuss] how to output multiple Kekule structures
>
> Hi Jim,
>
> The code currently has no way to enumerate Kekule structures. I don't
> recall this coming up in the past and, to be honest, it doesn't seem all
> that generally useful.
>
> Perhaps there's an alternate way to solve the problem; what are you trying
> to do?
>
> -greg
>
>
> On Mon, Sep 11, 2017 at 5:04 PM, James T. Metz via Rdkit-discuss <
> rdkit-discuss@lists.sourceforge.net> wrote:
>
> Hello,
>
>     Suppose I read in an aromatic SMILES e.g., for benzene
>
>     c1ccccc1
>
>     I would like to generate the major canonical resonance forms
> and save the results as two separate molecules.  Essentially
> I am trying to generate
>
>     m1 = 'C1=CC=CC-C1'
>     m2 = 'C1C=CC=CC1'
>
>     Can this be done in RDkit?  I have found a KEKULE_ALL
> option in the detailed documentation which seems to be what I
> am trying to do, but I don't understand how this option is to be used,
> or the proper syntax.
>
>     If it is necessary to somehow renumber the atoms and re-generate
> Kekule structures, that is OK.  Thank you.
>
>     Regards,
>     Jim Metz
>
>
>
>
>
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Rdkit-discuss mailing list
> Rdkit-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>
>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss

Reply via email to