Structures with Multi Centers

Hello,

 

I have a question regarding the correct indication of multi center bonds with Marvin when editing SDFs. Attached is an example. The compound is a mixture of 2 isomers (interconverting into each other) with the Methyl-anthraniloyl group being bound to any of the two Oxygens at the Ribose moiety. The other Oxygen will be bound to a Hydrogen atom (i.e. a Hydroxyl group).

 

Will my SDF show that correctly? I'm concerned because actually my structure still has two OH groups, while in fact only one is present in the molecule. However, I cannot get rid of the Hydrogen, Marvin always adds it to the Oxygen atom.

 

Thank you!

 

Best regards,

 

Thomas.

Hi Thomas,

 

Sorry for my late answer. The display of implicit hydrogens can be switched off in the View > Implicit Hydrogens menu.

Regarding the multicenter atom (we do not have multicenter "bond") my collegue will answer to you today.

 

Best regards,

 

Akos

Hi Thomas,

Unfortunately, the multicenter atom doesn't influence the automatic calculation of implicit hydrogens. You can set the valence of one of the oxygen atoms to 1 and then the hydrogen disappears. The chemical structure will be correct in this case.

 

Regards,

Erika.

Hello Akos + Erika,

 

thank you very much for your help!

 

Akos, I don't think that switching off the implicit hydrogens would help - this does only affect the view of the molecule, but the one excess hydrogen will still stay there, so as a consequence the molecular formula and mass will be wrong.

 

Erika, I edited the molecule according to the procedure you described, it's attached (NU-201.sdf). Will this reflect the chemical nature of this molecule correctly, or will NU-201_version2.sdf do that in a better way?

 

Thank you again, and have a good weekend!

 

Best wishes,

 

Thomas.

Hi,

It depends on what do you want to use it for. The second one is closest to the reality, but if you want to use this for searching, then I think the first one is better, since there at least one OH group remain, and you will find at least that isomer (while in the second case, if someone searches for the OH group, then will find neither one). Actually, is it very important to represent this isomerism in a single structure? If not, you could draw both structures (without muticenter), and a resonance arrow between them. Or -  especially for storing in a database - you could specify it as a mixture (and then it will be a hit in case of any SSS towards them), or if duplicate search is the major reason, then even if you simple include both isomers on the same record, the full fragment search of JChem will find either one.

Best regards,

Akos

Dear Akos,

thank you very much for the explaination. The SDFs will mainly be used for CAS number determination by CAS services. I already submitted a number of SDFs that they used successfully, but all of them were single-isomer compounds. So far I have not found a way to properly depict mixed-isomer ones so their database search will succeed. I was assuming that there is one and only correct way to do this in an SD file, but obviously there are multiple ways, so I guess we have to try them all out.

Since this isomerization problem applies to many of our products (most of them 2'/3'-substituted nucleotides), with some very important analogs among them, it's rather important for me to find a solution.

The SD files will very likely also be used for substructure search purposes later on, both in-house and on our website, but this is still some way down the road.

To explain this a little bit more in detail, find attached a number of SDFs. NU-201-1-sdf and NU-201-2.sdf contain individual isomers and were used successfully by CAS to determine CAS registry numbers:

ID   NU-201-1
REG  125902-32-1

ID   NU-201-2
REG  85287-55-4

I also made a SDF containing both structures in one file (NU-201-MIX.sdf), but this one failed.

I know that the isomer mixture has its own CAS number (151481-85-5), which is different from the numbers of the individual isomers, so there MUST be a way to depict this in an SDF which can then be used by CAS for database search.

I drew a mixture-containing SDF according to your description, it's also attached (NU-201-mixture.sdf). Do you think that this is correct?

Thank you!

Best wishes,

Thomas.

Hi Thomas,

Yes, I think the "NU-201-mixture.sdf" is the correct representation of this isomer mixture, but the question is whether CAS services supports storing and searching mixtures properly.

If they used JChem it would :) (but in that case even "NU-201-MIX.sdf" could be used).

Best regards,

Akos

Hi Akos,

 

I have submitted these variants to CAS and will see if any of these will work - I will post the results here.

 

Best wishes,

 

Thomas.