cyclopentadienide isomers

Even after aromatization, the following two molecules are not recognized as being identical to one another, because the negative charges are viewed as being localized on particular C atoms:

Code:

Marvin  01190521102D   8  9  0  0  0  0            999 V2000    -2.8750    1.5455    0.0000 C   0  5  0  0  0  0  0  0  0  0  0  0    -3.5424    1.0606    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -2.2076    1.0606    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -2.4625    0.2760    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -3.2875    0.2760    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -3.5424   -0.5086    0.0000 C   0  5  0  0  0  0  0  0  0  0  0  0    -2.8750   -0.9935    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -2.2076   -0.5086    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0   1  2  1  0  0  0  0   1  3  1  0  0  0  0   2  5  2  0  0  0  0   4  3  2  0  0  0  0   4  8  1  0  0  0  0   5  6  1  0  0  0  0   6  7  1  0  0  0  0   7  8  2  0  0  0  0   4  5  1  0  0  0  0 M  CHG  2   1  -1   6  -1 M  END

Code:

Marvin  01190521092D   8  9  0  0  0  0            999 V2000    -2.8750    1.5455    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -3.5424    1.0606    0.0000 C   0  5  0  0  0  0  0  0  0  0  0  0    -2.2076    1.0606    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -2.4625    0.2760    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -3.2875    0.2760    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -3.5424   -0.5086    0.0000 C   0  5  0  0  0  0  0  0  0  0  0  0    -2.8750   -0.9935    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0    -2.2076   -0.5086    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0   1  2  1  0  0  0  0   1  3  2  0  0  0  0   2  5  1  0  0  0  0   4  3  1  0  0  0  0   4  8  1  0  0  0  0   5  6  1  0  0  0  0   6  7  1  0  0  0  0   7  8  2  0  0  0  0   4  5  2  0  0  0  0 M  CHG  2   2  -1   6  -1 M  END

Can this be fixed?

This may be a tough one. (It is already on our TODO list for a while.)





We will think about how to represent these delocalized charges.

In ACE Organic (formerly EPOCH) I am able to confirm the identity of the two structures by comparing their sigma-bond networks and their charges, meaning they are resonance structures of one another. That works when you want different resonance structures to be treated as identical for any kind of compound (e.g., R₂C=O <-> R₂C⁺–O^–). Perhaps you could add such a flag to the options.

bobgr wrote:

In ACE Organic (formerly EPOCH) I am able to confirm the identity of the two structures by comparing their sigma-bond networks and their charges, meaning they are resonance structures of one another. That works when you want different resonance structures to be treated as identical for any kind of compound (e.g., R2C=O <-> R2C+–O–). Perhaps you could add such a flag to the options.

Thanks, we will consider this approach as well.

Here's another suggestion: When you aromatize a structure with n atoms in the ring and one of the atoms is charged, give each atom in the ring 1/n of the charge. In my example, the ring fusion atoms would have charges of -0.4, and the other atoms would each have a charge of -0.2.

It seems that we will be able to implement this in the search itself with no standardization step. For aromatic systems the total charge will be compared, and not at individual atoms.





Depending on the speed issues and other possible implications, it may be a separate search option.





Expect this functionality in a few months, but in the longer term we want to solve the same for any delocalised systems other than aromatic.