Generating stereoisomers with given chirality as constraint

Hello,





I'm using cxcalc with stereoisomers to generate stereoisomers for a compound that has a 2D SDF with a couple of chiralities included. But it looks like cxcalc does not respect the given chiralities and generates about 16 stereoisomers. I looked at the manual (pasted below) but couldn't find an option to let the program use the given chiralities as constraints. Anyone has suggestions?





Thanks,





Qing


------------------------------------





stereoisomers





Generates stereoisomers of the molecule.


Options:


-f, --format <output format> (default: sdf)


-m, --maxstereoisomers <maximum number of stereoisomers to be generated> (default: 200)


-v, --verify3d [true|false] if true invalid 3D structures of genereated stereoisomers are filtered


-3, --in3d [true|false] if true 3D structures are generated (invalid 3D structures are filtered)


Example:





cxcalc stereoisomers -v true test.sdf

Hi Qing,





You are right, the steroisomer plugin can not use the given chiralities as constraints. We will add this feature in a later release (we scheduled it for Marvin 5.2).





Best regards,


Zsolt

Hello, are there any news regarding this topic?





Best regards, Ingo

Hi,





We postponed the development of this feature. Would this feature be important / urgent for you?





Zsolt

Hi,








I am using the stereoisomers plugin to fill my compound database with entries from different sources/quality. Therefore it is of great importance for me to use as much information from the input structure as possible. If a structure is given with chirality I don't want to enumerate all isomers. In contrast to that, a structure with no stereoinformation should be handled as racemate.





Calculating stereoisomers with constraints on stereocenters with given chirality would be a great help for me. Often, the calculation of tautomers leads to new stereocenters. In that case the calculation of the isomers should not alter stereoinformation on other stereocenters.





Are you still planning to implement this feature in the forseeable future?





Best Regards, Ingo

Quote:

Are you still planning to implement this feature in the forseeable future?

Yes, we plan to add this feature to Marvin 5.2.3. 





Best regards,


Zsolt

All you need to do to add this feature is to do a series of JChem exact matches (MolSearch) with the original compound as the query and each generated stereoisomer as the targets.  

Every stereoisomer that has a specified configuration where the original compound has either an unspecified configuration or the same configuration will match.  Save the matches and discard the rest.  

bobgr wrote:

All you need to do to add this feature is to do a series of JChem exact matches (MolSearch) with the original compound as the query and each generated stereoisomer as the targets. Every stereoisomer that has a specified configuration where the original compound has either an unspecified configuration or the same configuration will match.  Save the matches and discard the rest.

Bob, is it a workaround you suggest to other users?

Zsolt

Yes, that's why I posted it.  

It's not really a workaround, because there's no bug in the plugin.  Just some additional code to enhance an existing feature.

But I do think you should still develop the method yourselves, because it's faster to generate, say, four out of the sixteen possible stereoisomers than it is to generate all sixteen and then cast out twelve.

OK, I see.

We will develop it.

Zsolt

Zsolt wrote:

Quote: Are you still planning to implement this feature in the forseeable future? Yes, we plan to add this feature to Marvin 5.2.3.  Best regards, Zsolt

 

I just read the release notes of Marvin 5.2.3 and could not find any news about this feature. I'm still looking forward to it, as it would cause an enormous speed or the isomer calculation.

 

Did you postpone it again?

 

Greets,

Ingo

It's implemented, and included in Marvin 5.2.3 release.

See:

http://www.chemaxon.com/marvin/help/calculations/isomers.html#stereoisomer

http://www.chemaxon.com/marvin/help/applications/cxcalc-calculations.html#stereoisomers

It seems we forgot to include this feature in the changes list. Sorry for that, we will include it on the next update.

Regards,

Zsolt

Great, thanks a lot.

 

Edit: I think is does not work as intended. Using the Sketch interface, all isomers are generated even if the chirality is specified.

I can confirm that the new restrictions work in the API.

            if (excludeSpecified) { // new to JChem 5.2.3

                plugin.setProtectTetrahedralStereo(true);

                plugin.setProtectDoubleBondStereo(true);

            }

It seems to work on the command line, but not using the GUI...

The result is not correct if the Generate:both option is set. We will fix this in the next release, sorry for the inconvenience.

Zsolt

Some correction: the bug only appears if the structure is drawn manually. If it is imported or copied to MarvinSketch, then the StereoisomerPlugin works correctly. Meanwile we fixed the bug causing this issue, the fix will be available in 5.2.4.

Zsolt