Incorrect Chiral Center Counts

I am using MarvinSketch 6.0.4.  I noticed that that Topology analysis gives the incorrect Chiral Center Count for irofulven :  CC1=C(CO)C2=C(C)C3(CC3)[C@@](C)(O)C(=O)C2=C1    This structure has one asymmetric atom and one chiral center, but the Topology Analysis give a chiral center count of 2.  The simplified irofulven substructure O[C@H]1CCCCC11CC1 shows the same problem.

 

Hi,

We can confirm that this is a bug in our parity calculation. We will fix it in Marvin 6.2. Is it Ok for you?

Regards, Domi

Hi, how do you differentiate an "asymmetric atom" from a "chiral center?" Those terms are usually used interchangeably.  (Actually, "chiral center" is discouraged; we are supposed to use "stereogenic center.")  

According to the documentation:

• Asymmetric atom count: the number of asymmetric atoms (having four different ligands).


• Chiral center count: the number of tetrahedral stereogenic centers. This function identifies two chiral centers in 1,4-dimethylcyclohexane, which does not contain asymmetric atoms.

Your distinction is relies on a misunderstanding.  1,4-Dimethylcyclohexane does indeed have two atoms that have four different ligands.  On each such atom, one of the ligands resides in an R environment, and the other in an S environment, as determined by the other stereocenter in the molecule.

For tetrahedral atoms, "has four different ligands" and "is a stereocenter" are required to be exactly synonymous by mathematics.  (We can consider a lone pair as a ligand.)  

You should deprecate the terms "asymmetric atom" and "chiral center" and simply use the well-defined, universally accepted "stereogenic center" or "stereocenter" for short.  The term "chiral" refers to the symmetry of entire molecules, not to individual atoms.  For example, there is nothing "chiral" about the atoms in 1,4-dimethylcyclohexane, both of whose diastereomers are achiral compounds.  

Dear Bob,

Thank you for the clarification.

"You should deprecate the terms "asymmetric atom" and "chiral center" and simply use the well-defined, universally accepted "stereogenic center" or "stereocenter" for short. "

Yes, we will do this and may try to help our users to differentiate stereogenic centers by type.

Something like tetrahedral, axial, cis-trans. Miklos will add his comment to this topic what is our plan here.

Dear Bob and Tim,

Based on your feedbacks we have a proposal to improve the Geometry/Topology Analyzer Calculator Plugin for Stereochemistry related calculations.

As Bob pointed out the name "Chiral Center Count" is somewhat misleading. Therefore, we want to remove this one as a calculator.

We propose to introduce the calculation for "stereogenic center count", which is in line with the IUPAC notation.

We also propose to introduce calculators for various subgroups as it follows:

"asymmetric atom count" - The number of asymmetric atoms having four different ligands. (This would remain as it is now in the calc. plugin.)

"cis-trans stereogenic center count" - The number of cis-trans stereogenic centers.

"pseudo asymmetric atom count" - The number of pseudo asymmetric atoms.

"atrop stereogenic center count" - The number of centers causing atrop isomerism.

Please let us know you thoughts on this proposal.

 

Best Regards,

Miklos

mjszabo wrote:

We propose to introduce the calculation for "stereogenic center count", which is in line with the IUPAC notation. We also propose to introduce calculators for various subgroups as it follows: "asymmetric atom count" - The number of asymmetric atoms having four different ligands. (This would remain as it is now in the calc. plugin.) "cis-trans stereogenic center count" - The number of cis-trans stereogenic centers. "pseudo asymmetric atom count" - The number of pseudo asymmetric atoms. "atrop stereogenic center count" - The number of centers causing atrop isomerism.

Hello,

A number of questions and further suggestions:

1. Will "stereogenic center count" include stereogenic trigonal, trigonal bipyramidal, and octahedral stereocenters?  If so, good, but you need to make clear that it applies to all geometries.  If not, then this value should be renamed "tetrahedral stereocenter count", and its value is exactly the same as "asymmetric atom count", as indicated above.  (In other words, "asymmetric atom count" would still be superfluous.)


2. Does "cis-trans stereogenic center count" include both tetrahedral stereocenters on rings and trigonal stereocenters in double bonds?  If only the former, I suggest naming it "tetrahedral cis-trans stereocenter count."


3. I assume, by "pseudoasymmetric atom," you mean a stereogenic tetrahedral atom whose two stereoisomers are both achiral, as in (2R,3r,4S)- and (2R,3s,4S)-2,3,4-pentanetriol.  


4. By "atrop stereogenic center count," are you meaning to include compounds that are chiral by virtue of restricted rotation, such as 2,2´-binaphthol, in addition to such atoms as the central ones in odd-number cumulenes such as 2,3-pentadiene?  If not, perhaps you should name it, "stereogenic cumulene count."  (I should point out that the terminal atoms, not the central ones, in odd-number cumulenes are stereogenic, but we often refer to the cumulene as a whole as a stereogenic group.)  


5. What about compounds such as η⁵-(1-ethyl-2-methylcyclopentadienyl)-η⁵-cyclopentadienyliron, often called planar-chiral compounds?

I ♥ stereochemistry.

-- Bob

Dear Bob,

thank you for your prompt answer.

1. We intend to cover all subgroups if possible. If we can't do it on short term, we'll list them as known limitation.

2. For the cis-trans we may cover two subgroups as you wrote

"cis-trans tetrahedral stereocenter count" for tetrahedral cases

"cis-trans trigonal stereocenter count" for double bonds.

3. Yes, your description and example is good: "pseudoasymmetric atom," is a stereogenic tetrahedral atom whose two stereoisomers are both achiral, as in (2R,3r,4S)- and (2R,3s,4S)-2,3,4-pentanetriol.

4. Yes, the atrop isomer subgroup might be refined as you write. We intend to cover all subgroups. If we can't do it on short term, we'll list them as known limitation.

5. We haven't thought of this one, but we'll try our best to solve it.

 

Thank you very much for your valuable feedbacks,

Miklos