On Tue, Mar 31, 2015 at 7:36 PM, Soren Wacker <swac...@ucalgary.ca> wrote:
>
> thanks. Is the random conformation really randomly selected, or is it
> always the same? I mean, if I generate a conformation from smiles without
> chiral specifiaction, will I end up always with the same conformation?
>
>
The example below demonstrates that you can get different stereochemistry
in the conformers.
If you provide a random number seed for the generation of a single
conformer, you should always get the "same" conformer (to within roundoff
error).
> So, I am using the Chem.Lipinski and Chem.Descriptors modules to calculate
> properties of molecule objects that where generated by smiles strings. Are
> all descriptors independent of the particular conformations? Or are you
> aware of descriptors which definitely depend on a conformation?
>
I believe that all of the RDKit 2D descriptors (the ones in
Chem.Descriptors, which include Chem.Lipinski) are not sensitive to
chirality. You could prove this to yourself by providing a number of
molecules with and without stereochemistry in the SMILES and check to see
if you get different answers. Note that since these descriptors are all
"2D" (=determined by the connectivity of the molecule), none of them are
sensitive to conformation.
I would like to generate descriptors that are reproducible for a given
> compound, ideally independent of the conformation. One of the problem is
> that sometimes smiles strings contain chiral specification, sometimes
> different protonation states etc. I need to standardize it somehow.
>
The protonation state problem is more difficult. Many of the 2D descriptors
certainly are sensitive to charge and number of Hs.
There is some sample code for neutralizing molecules in the RDKit cookbook:
http://www.rdkit.org/docs/Cookbook.html
-greg
------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
