Marvin Sketch keeps changing my molecule's stereochemistry!

Hi all, I apologise in advance for any mistakes I make in forum etiquette as this is my first post here.

I'm trying to use Marvin Sketch to input molecules into Instant JChem. I am using external databases such as ChemSpider and RCSB PDB to check IUPAC names. Then I copy & paste the IUPAC name of each molecule into Marvin Sketch, one at a time.

I am dealing with ligands binding to proteins, so the stereochemistry of the molecule is important.

Right now I have two ligands that need to be kept separate since they have different stereochemistry:

(4E,6Z,8S,9S,10E,12S,13R,14S,16S)-13-hydroxy-8,14,19-trimethoxy-4,10,12,16-tetramethyl-3,20,22-trioxo-2-azabicyclo[16.3.1]docosa-1(21),4,6,10,18-pentaen-9-yl
carbamate [PDB Ligand ID = GMY]

and

(4E,6Z,8S,9S,10E,12S,13R,14S,16R)-13-hydroxy-8,14,19-trimethoxy-4,10,12,16-tetramethyl-3,20,22-trioxo-2-azabicyclo[16.3.1]docosa-1(21),4,6,10,18-pentaen-9-yl
carbamate [PDB Ligand ID = GDM]

I have Instant JChem set with the options Duplicate filtering OFF and Assume absolute stereochemistry ON.

Two problems are occurring:

 

1) Marvin Sketch is reversing the stereochemistry of my molecule. When I paste isomer GDM into the Structure --> Edit molecule --> Input name field, Marvin Sketch is outputting (8R,9S,12S,13R,14S,16R)-13-hydroxy-8,14,19-trimethoxy-4,10,12,16-tetramethyl-3,20,22-trioxo-2-azabicyclo[16.3.1]docosa-1(21),4,6,10,18-pentaen-9-yl carbamate

i.e. the chiral centre on carbon 8 is being inverted for no apparent reason.

 

2) If I then go back into that Edit molecule --> Input name field menu, and force the molecule to be 8S, Marvin Sketch flips around my double bonds! I end up with a structure that I don't understand. See the attached screenshot. I don't know what this means since I thought wavy lines meant "either" stereochemistry at a tetrahedral carbon, but this is happening to a section of the molecule between two trigonal planar carbons!

Also, as you can see, Marvin Sketch then decides it doesn't like the geometry, and that the carbamate group coming off carbon 9 must have a steric clash with the methyl group on carbon 10!

 

Does anyone have any idea how to fix this?

I'm using Instant JChem and Marvin Sketch 5.7.0, the most recent version - though this same problem occurred with the previous version too.

Thank you for your detailed report. As you have correctly identified, there are two separate issues here.

The first is that indeed, the chirality of carbon atom 8 is wrongly imported. I have identified the reason for this bug, and I'm planning to fix it in time for the 5.9 release.

The second bug is that the 4E,6Z, and 10E descriptors are missing in the generated name. (This is the reason why marvin reimports that generated name with "unknown" stereoconfirmation for the double bonds.) I have just fixed that bug, it will therefore surely be released in 5.9.

Until these fixes are released, if you have only one or a few structures to import, the way to correctly fix the structure manually is to right click on the atom with the wrong chirality, and set it with the Stereo -> R/S contextual menu. If this is a problem, we can discuss sharing a beta version earlier with those fixes.

Best regards,

Daniel

By the way, you can display stereochemistry information on the structure itself using the View -> Stereo menu.

baratron wrote:

Daniel Bonniot wrote:

Thank you for your detailed report. As you have correctly identified, there are two separate issues here. The first is that indeed, the chirality of carbon atom 8 is wrongly imported. I have identified the reason for this bug, and I'm planning to fix it in time for the 5.9 release. The second bug is that the 4E,6Z, and 10E descriptors are missing in the generated name. (This is the reason why marvin reimports that generated name with "unknown" stereoconfirmation for the double bonds.) I have just fixed that bug, it will therefore surely be released in 5.9. Until these fixes are released, if you have only one or a few structures to import, the way to correctly fix the structure manually is to right click on the atom with the wrong chirality, and set it with the Stereo -> R/S contextual menu. If this is a problem, we can discuss sharing a beta version earlier with those fixes. Best regards, Daniel

Hi Daniel,

Thanks for your reply. I'm glad that both problems are fixable bugs rather than something weird at my end.

 

There is more to the problem at carbon 8. I have tried changing the absolute configuration using the Stereo --> R/S menu, and then Marvin imports the correct structure into Instant JChem. However, it does not change the name! So then I have a database where the IUPAC name does not match the structure (please see attached screenshot).

 

I have about 130 structures of this type to import, so I would appreciate having the fixes now if that is possible. 

 

The biggest problem from my perspective is that my College deadline for the work I'm doing is this Thursday! *STRESS* If you could tell me when those fixes might be ready, I might be able to get an extension based on the fact it's not my fault that there were bugs in the program.

Helen-Louise

 

P.S. paitocalva - neither of your posts have any content. I don't know if you're trying to say that you have the same problem as me, or offer a fix, or ...?

Hi Helen-Louise,

I could share with you an alpha version which fixed the name generation bug (missing E and Z), if that's useful. For the structure generation bug (wrong chirality), I hope I can have it during next week, but I cannot guarantee it.

I will ask somebody to answer you about the IJC problem (name not regenerated).

Regards,

Daniel

The IUPAC names in IJC should be automatically updated when chemical structure is changed, if the field is dynamically generated by using Chemical Terms. Therefore this depends on the nature of the IUPAC name field as well. There are basically two possibilities:

1. IUPAC name field is just a text-field and it's value is fixed (imported or manually added - no automatic change is possible besides by using scripts, plugins).

2. IUPAC name field is Chemical Terms field and the name always follows the change in the structure.

I recommend you to use the seccond option as it's easy to add new CT field and you don't need to worry about it any more. You probably know it, but anyway here is a link to docs about CT fields: https://www.chemaxon.com/instantjchem/ijc_latest/docs/user/help/htmlfiles/editing_database/fields_chem_terms.html

Good luck

Filip

Daniel Bonniot wrote:

I could share with you an alpha version which fixed the name generation bug (missing E and Z), if that's useful. For the structure generation bug (wrong chirality), I hope I can have it during next week, but I cannot guarantee it.

Hi Daniel,

Alpha versions sound scary! I think I'll wait until it's at least in beta. For the work I have to hand in tomorrow, I shall simply copy & paste the diagrams from academic papers (with references) and point my assessors at this thread to show that there is a genuine problem.

 

fzimandl wrote:

The IUPAC names in IJC should be automatically updated when chemical structure is changed, if the field is dynamically generated by using Chemical Terms. Therefore this depends on the nature of the IUPAC name field as well. There are basically two possibilities: 1. IUPAC name field is just a text-field and it's value is fixed (imported or manually added - no automatic change is possible besides by using scripts, plugins). 2. IUPAC name field is Chemical Terms field and the name always follows the change in the structure.

Hi Filip,

Well, that's the weird thing - I am using the IUPAC Chemical Terms field! That's why I'm so surprised that it isn't being updated!

The steps that I've gone through are:

1) Add a new row to my database in Instant JChem. Structure is a compulsory field in my database, so it's the first box I need to fill in.

2) Typing (8S,9S,12S,13R,14S,16R)-13-hydroxy-8,14,19-trimethoxy-4,10,12,16-tetramethyl-3,20,22-trioxo-2-azabicyclo[16.3.1]docosa-1(21),4,6,10,18-pentaen-9-yl
carbamate into the Marvin Sketch window that appears inside the Structure field. It produces the incorrect isomer, 8R.

2) Highlighting carbon 8 and going to the Stereo --> R/S menu to change its chirality to S as it should be.

3) The new structure appears in the Instant JChem window, but the IUPAC Name field doesn't change.

You could try reproducing this at your end with version 5.7 of Marvin Sketch and Instant JChem? If your version of the program works differently, please let me know and I can zip up a copy of my database and send it to you.

I am using the Mac version, in case that makes any difference.

Thanks for all your help!

Helen-Louise.

Hi Helen-Louise,

if I understand correctly the highlighted atom in the picture is actually at the possition 9 and it seems that change is being propagated to the IUPAC name after saving. There is methyl group in the picture at the possition 8(you mentioned methoxy, which changes group priority) - try to change stereochemistry there.

Ragarding CT field updates, I think it works properly(at least on my computer, IJC 5.7, Windows OS).

You can send the project to my email if the problem still persist(honestly I'm not sure where could be the difference) and I'll try it ones again or we can test it on Mac.

Regards,

Filip

Hi Helen-Louise,

Both naming bugs (missing E and Z, and 8R instead of 8S) are now fixed and will be released in version 5.9. I can share with you an alpha version anytime if you need.

Cheers,

Daniel