Combinatorial substitutions for two R-groups

Dear users and moderators,

I want to use Jchem or Marvin to make all possible ligands for a structure with two R-groups. Say, the core is disubstituded benzene (R1-C6H4-R2) and there are N1 radicals for R1 group and N2 radicals for R2 group. So, I want to get N1*N2 molecules (approx. 400) with 3D coordinates, ready for docking.

What should I do, starting from drawing structures of the core and R1 and R2 radicals? Is it a subject of molconvert with -R<i> option, or Fragmenter, or Reactor?

Thanks in advance,

Anton.

Hi Anton,

 

Yes, you should draw a Markush structure that includes the scaffold and the R-group definitions.

Then you can use Markush enumeration to generate the specific structures.

 

Molconvert can then create 3D coordinates using the -3 option.

 

See more details here:

http://www.chemaxon.com/library/drawing-markush-structures-with-r-groups/

https://www.chemaxon.com/marvin/help/calculations/markush.html

 

Best regards,

Szabolcs

Thanks, Szabolcs. I was managed to do that with:

cxcalc markushenumerations -s true -o true markush.mrv -f sdf -i reagent > ligands.sdf

molconvert mol2:H ligands.sdf -3:"S{fine}" -o ligands.mol2

But there's a problem: in SDF I have a line ">  <Markush code> // reagent(R2(18):b01, R1(19):a01)", where b01 and a01 are R-groups' members IDs, that I could understand what's the molecule it is, which substituent groups does it have. But this information is lost when I save in mol2 format, which is better for my further tasks. How I can save these "reactant" IDs in mol2, and maybe also in separate mol2 files' names?

Next, during the enumeration and/or conversion sometimes atom names in final mol2 files is not the same in the core region. Although, I need them to be the same -- is is for further RMSD calculations after docking.

Are there general ways to overcome these difficulties?

Anton.

 

Hi Anton,

 

The Markush ID-s are stored in a field in case of SDF output (also called molecule property). I don't think that this information can be represented in the mol2 file format.

 

We do not have a ready utility to separate the SDF into different files and make one property the file name, but you may want to use our Java or .NET API to write such a program yourself.

You will need to use the following classes and methods:

http://www.chemaxon.com/marvin/help/developer/beans/api/chemaxon/formats/MolImporter.html

http://www.chemaxon.com/marvin/help/developer/beans/api/chemaxon/formats/MolExporter.html

http://www.chemaxon.com/marvin/help/developer/beans/api/chemaxon/struc/Molecule.html#getPropertyObject%28java.lang.String%29

 

Regarding the second question: Indeed, the enumeration removes the molecule name field during enumeration.

However, you can use the "ID tag name" to include in the Markush code.

See more details:

http://www.chemaxon.com/marvin/help/calculations/markush.html#function_markushcode

 

Best regards,

Szabolcs