How to convert Inchi fomat to 3D format

Dear all

Can anybody tell me how to convert a set of inchi files to 3D files (mol2 or sdf or pdb) without loosing stereochemical information. I tried molconvert to to perform, however, it retained the stereochemical infomation, but the 3D coordiantes are not all correct.

I am a new user of molconvert, it would be great if anybody tell me what is the best sysntax to convert either ichi or inchikey to 3D format.

Thank you

Vinod

Dear Vinod,

We have used the inchi executable to convert inchi to mol till the current version (5.4.1), but from 0d input  (InChI without AuxInfo) it converts the structure into a 0d molfile, where no stereo information is present. The upcoming 5.5 version will use the IUPAC InChI API and will be able to convert inchis to 3d structures, with correct stereo, You will be able to use it from molconvert

molcovert -3 mol inchifilename.inchi

InChIKey can not be convertted to structure, it is something like a hash.

Best Regards,

Peter

Hi Vinod,

You also metioned, that "the 3D coordiantes are not all correct." 

Can you give an example? or explain the problem more specifically?

 

Thanks

Miklos

Hi Mvargyas

Attached are the three different file formats for a specific compound which were converted using molconvert version 5.3.1.

The actual 3D file can be visualized here: http://www.drugbank.ca/drugs/DB00560/structure?dim=3d

I have converted ~1000 inchis to 3D formats, only some of them were wrong. please look at the attachments, specifically the CO-NH2 at both ends of the compound.This is where the compound is wrong, otherwise the conversion is perfect with stereochemistry included.

The back bone of the compound was converted correctly, please see the attached PDB with 'noH'.

I used the following syntax for the conversion:

molconvert pdb -3 DB00560.sdf.inchi -o DB00560_noH.pdb (without hydrogen)

molconvert mol2:H -3 DB00560.sdf.inchi -o DB00560_mol2.mol2

Any help would be appreciated.

Thank you,

Vinod Kasam

 

Hi Vinod,

I cannot find the "CO-NH2 part " of the molecule "DB00560_sdf.sdf". Do you think of the orientation of the hydrogens in the imidic acid - like moiety -C(OH)=NH ? This moiety is not planar in the version you tried but will be correctly planar in the latest Marvin version: 5.4.1.1 . 

The underlying forcefield has been improved a lot since 5.3.1. In the new version you can also access the industry standard MMFF94 forcefield from the MarvinSketch gui under the Tools\Conformation\Conformers dialog box. In this box select the "Calculate lowest energy conformer" and the "MMFF94 optimization"

Is this helps?

Cheers,

Adrian