Re: [Rdkit-discuss] Error in RDKit output for finding ring atoms!

2021-03-11 Thread Ling Chan
Hello Goutam,
There are not even 30 atoms in your smiles string. This is because, for
example, [C@H] denotes one carbon atom. The H is there to describe the C.
It is not an atom in the smiles string.
Ling


Goutam Mukherjee  於 2021年3月11日週四 上午8:35寫道:

> Dear Members,
>
> I have found an error in RDKit output. I am not sure whether it is my
> mistake.
> I have a SMILES code of  a molecule:
> C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@
> @H]1O)N1C=NC2=C1N=CN=C2N
>
> the 3D coordinates of the molecule is attached here with.
>
> *I ran the following command:*
>
>
>
>
>
>
>
> *In [1]: from rdkit import ChemIn [2]: m =
> Chem.MolFromSmiles('C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@@H]1O)N1C=NC2=C1N=CN=C2N')In
> [3]: ri = m.GetRingInfo()In [4]: print(ri.AtomRings())((10, 11, 12, 13,
> 15), (18, 17, 21, 20, 19), (22, 23, 24, 25, 20, 21))*
>
> Here the atom rank does not correspond to the ring atom ranks.
> This molecule contains two five members and one six member ring
> The true atom rank would be
> *{(11, 13, 14, 16, 19), (22, 23, 24, 25, 26), (25, 26, 27, 28, 29, 30)}*
>
> Could anyone please give me a solution how I get a corect atom ranks which
> are part of a ring.
>
> Thanks and Best Regards,
> Goutam
>
> ___
> Rdkit-discuss mailing list
> Rdkit-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>
___
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss


Re: [Rdkit-discuss] Error in RDKit output for finding ring atoms!

2021-03-11 Thread Goutam Mukherjee
Dear Ivan,

Many thanks for your reply.
Yes, when I give 3D hydrogen added molecule as input (molecule.pdb) the
result remains the same.
Say, if I give benzene molecule (C1=CC=CC=C1) as an input, or other complex
molecule, it prints the correct atom rank no matter whether hydrogen atoms
are there or not.
In [1]:  from rdkit import Chem
In [2]:  m = Chem.MolFromSmiles('C1=CC=CC=C1')
In [3]: ri = m.GetRingInfo()
In [4]:  print(ri.AtomRings())
*((0, 5, 4, 3, 2, 1),)*


Thank and Best Regards,
Goutam


On Thu, Mar 11, 2021 at 5:49 PM Ivan Tubert-Brohman <
ivan.tubert-broh...@schrodinger.com> wrote:

> Hi Goutam,
>
> The ring atoms reported by RDKit in your example are correct; you just
> need to consider that the atom indexes correspond to the position of each
> atom in the SMILES string. How could RDKit guess the index that the atom
> might have in a PDB file that's not even being read in your example?
>
> I'm guessing maybe in your real use case you did read the PDB file. It is
> possible that the atoms got renumbered, for example if the hydrogens were
> deleted in the process.
>
> Hope this helps,
> Ivan
>
> On Thu, Mar 11, 2021 at 11:35 AM Goutam Mukherjee 
> wrote:
>
>> Dear Members,
>>
>> I have found an error in RDKit output. I am not sure whether it is my
>> mistake.
>> I have a SMILES code of  a molecule:
>> C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@
>> @H]1O)N1C=NC2=C1N=CN=C2N
>>
>> the 3D coordinates of the molecule is attached here with.
>>
>> *I ran the following command:*
>>
>>
>>
>>
>>
>>
>>
>> *In [1]: from rdkit import ChemIn [2]: m =
>> Chem.MolFromSmiles('C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@@H]1O)N1C=NC2=C1N=CN=C2N')In
>> [3]: ri = m.GetRingInfo()In [4]: print(ri.AtomRings())((10, 11, 12, 13,
>> 15), (18, 17, 21, 20, 19), (22, 23, 24, 25, 20, 21))*
>>
>> Here the atom rank does not correspond to the ring atom ranks.
>> This molecule contains two five members and one six member ring
>> The true atom rank would be
>> *{(11, 13, 14, 16, 19), (22, 23, 24, 25, 26), (25, 26, 27, 28, 29, 30)}*
>>
>> Could anyone please give me a solution how I get a corect atom ranks
>> which are part of a ring.
>>
>> Thanks and Best Regards,
>> Goutam
>>
>> ___
>> Rdkit-discuss mailing list
>> Rdkit-discuss@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>>
>
___
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss


Re: [Rdkit-discuss] Error in RDKit output for finding ring atoms!

2021-03-11 Thread Ivan Tubert-Brohman
Hi Goutam,

The ring atoms reported by RDKit in your example are correct; you just need
to consider that the atom indexes correspond to the position of each atom
in the SMILES string. How could RDKit guess the index that the atom might
have in a PDB file that's not even being read in your example?

I'm guessing maybe in your real use case you did read the PDB file. It is
possible that the atoms got renumbered, for example if the hydrogens were
deleted in the process.

Hope this helps,
Ivan

On Thu, Mar 11, 2021 at 11:35 AM Goutam Mukherjee 
wrote:

> Dear Members,
>
> I have found an error in RDKit output. I am not sure whether it is my
> mistake.
> I have a SMILES code of  a molecule:
> C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@
> @H]1O)N1C=NC2=C1N=CN=C2N
>
> the 3D coordinates of the molecule is attached here with.
>
> *I ran the following command:*
>
>
>
>
>
>
>
> *In [1]: from rdkit import ChemIn [2]: m =
> Chem.MolFromSmiles('C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@@H]1O)N1C=NC2=C1N=CN=C2N')In
> [3]: ri = m.GetRingInfo()In [4]: print(ri.AtomRings())((10, 11, 12, 13,
> 15), (18, 17, 21, 20, 19), (22, 23, 24, 25, 20, 21))*
>
> Here the atom rank does not correspond to the ring atom ranks.
> This molecule contains two five members and one six member ring
> The true atom rank would be
> *{(11, 13, 14, 16, 19), (22, 23, 24, 25, 26), (25, 26, 27, 28, 29, 30)}*
>
> Could anyone please give me a solution how I get a corect atom ranks which
> are part of a ring.
>
> Thanks and Best Regards,
> Goutam
>
> ___
> Rdkit-discuss mailing list
> Rdkit-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
>
___
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss


[Rdkit-discuss] Error in RDKit output for finding ring atoms!

2021-03-11 Thread Goutam Mukherjee
Dear Members,

I have found an error in RDKit output. I am not sure whether it is my
mistake.
I have a SMILES code of  a molecule:
C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@@H]1O)N1C=NC2=C1N=CN=C2N

the 3D coordinates of the molecule is attached here with.

*I ran the following command:*







*In [1]: from rdkit import ChemIn [2]: m =
Chem.MolFromSmiles('C[S+](CC[C@H](N)C([O-])=O)C[C@H]1O[C@H]([C@H](O)[C@@H]1O)N1C=NC2=C1N=CN=C2N')In
[3]: ri = m.GetRingInfo()In [4]: print(ri.AtomRings())((10, 11, 12, 13,
15), (18, 17, 21, 20, 19), (22, 23, 24, 25, 20, 21))*

Here the atom rank does not correspond to the ring atom ranks.
This molecule contains two five members and one six member ring
The true atom rank would be
*{(11, 13, 14, 16, 19), (22, 23, 24, 25, 26), (25, 26, 27, 28, 29, 30)}*

Could anyone please give me a solution how I get a corect atom ranks which
are part of a ring.

Thanks and Best Regards,
Goutam


molecule.pdb
Description: Binary data
___
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss